“Stop the Spread”: A Comparison of Online Misinformation on the Coronavirus in
Different Western Countries
L.M.F.T. Van Kessel s1352024
Msc Crisis and Security Management Dr. T. Van Steen and Dr. E. de Busser
Abstract
Since January 2020, the world is in the grip of the coronavirus pandemic. Besides the health risks, the pandemic also caused a new spike in misinformation stories. Misinformation can have serious consequences, both societal and individual. It can influence people’s decisions and actions (Ruokolainen and Widen, 2019). Besides, misinformation can make people distrust certain information (Karlova and Fisher, 2013), their government (Shin et al., 2018) and services provided by the government (Lewandowsky et al., 2017). Therefore, the increase in misleading news on the coronavirus can have consequences for the preventative measures that are
encouraged by governments, or for the effect of health measures implemented by the World Health Organization.
This thesis aims to analyze the content of misinformation related to the coronavirus pandemic in several western countries. On top of that it makes a comparison between
misinformation in different western countries. Therefore the content of online misinformation on the coronavirus will be analyzed through a content analysis. Humprecht, Esser and Van Aelst presented a framework to compare western countries, based on their resilience to online
misinformation (2020). By comparing content from the Netherlands, the United Kingdom and the United States we might learn more about what type of misinformation stories are currently
competing with the mainstream media, and to what extent there is a difference in misinformation between these different countries.
There are notable differences in online misinformation between the Netherlands, the United Kingdom and the United States. In the United Kingdom and the Netherlands most of the online misinformation consisted of health related topics, and was spread by social media
accounts. In the United States, the country with the lowest resilience to online misinformation (Humprecht et al., 2020), most of the online misinformation was about politics and spread by their national government. These results might suggest that the coronavirus pandemic has turned into a political issue in the United States. It is highly unlikely that the health related online (conspiracy) misinformation isn’t spread in the United States. Since this research is limited to the coverage of fact check sites, future research could provide better insight in the current state of misinformation on the coronavirus, by analyzing a broader set of online misinformation directly from social media platforms and (conspiracy) websites.
Keywords: Fake News, Disinformation, Misinformation, Coronavirus, United States, The
Index
Introduction………4
Societal relevance……….6
Academic relevance………..6
Chapter 1: Literature review……….7
1.1 Misinformation, disinformation and “fake news”….……….……….7
1.1.1. Consequences of online misinformation………...10
1.2 Populism and misinformation..……….…………10
1.3 Misinformation in the United States..………..………….12
1.4 Misinformation in Europe...………..13
Chapter 2: Theoretical framework……….………….14
2.1 Defining fake news and misinformation………..……….………14
2.2 Defining resilience………16
Chapter 3: Method……….………...18
3.1 Study setting and sample selection……….…………...19
3.2 Measurements………...20
Chapter 4: Analysis……….……..22
4.1 Sample description……….…...22
4.2 Cross-national analysis……….22
4.2.1 Sources of online misinformation………...22
4.2.2 Topics of online misinformation...……….…23
4.2.3 Targets of online misinformation………...……….……...24
4.2.4 Type of online misinformation……...……….…...25
4.2.5 Type of misinformation and source of misinformation………...………...26
Chapter 5: Discussion………...27 5.1 Limitations………29 5.2 Recommendations……….……30 Chapter 6: Conclusion………..31 References………..32 Appendix 1: Codebook……….38
The risk of misinformation around Covid-19
Currently the world is in the grip of the coronavirus pandemic. Countries are in lockdown to prevent the spread of the virus, and to prevent unnecessary deaths. Governments, scientists and health organizations try to improve testing methods, to quickly identify infected citizens. According to Jon Rappoport’s online articles, what they do not tell you is that those test are inherently fraud and they force people into toxic treatments they do not need. The vaccine that researchers are currently developing will either become a DNA vaccine, which alters people's genetic makeup permanently, or a RNA vaccine, which attacks the autoimmune system. To be clear, corporations and governments are killing people, and they try to cover up their crimes with this virus. We need to remember that death and illness can be caused by multiple sources, but suddenly we are supposed to believe that they are all caused by this one virus (Rappoport, 2020). This information is shared by Jon Rappoport on his blog on the website
‘Nomorefakenews.com’. He is described as an American investigative journalist, and his articles are widely shared and praised, as one quick google search will show. Conspiracy theories are of course nothing new. In our modern western society we have been confronted with multiple conspiracies. These conspiracies warn us about the “hidden truth” behind for instance: HIV/AIDS, vaccinations, the deaths of JFK and princess Diana and the attacks on 9/11
(Harambam and Aupers, 2017). The average citizen probably stumbles upon a conspiracy or fake news article once in a while. Probably while scrolling through social media. But it seems that the current pandemic has opened the gates for those alternative theories. The uncertainty about the origin and characteristics of this news coronavirus seems to provide a platform for conspiracy websites, to present their theories to a wider audience. For instance, theories on the effects of 3G, 4G and 5G have existed for a while, but since someone linked 5G radiation to the current
pandemic, the 5G conspiracy gained a lot of attention.
The 5G conspiracy might be one of the most noticeable theories about the coronavirus. One of the first stories claimed that it was no coincidence that the coronavirus spread from Wuhan, where 5G technology was trialed (Tuters and Knight, 2020). Others claim that the virus was created, to make sure people stay indoors, while the 5G network gets installed everywhere. Some claim that the radiation weakens the immune system, which makes people more susceptible for the virus, but there are also people who believe that the virus is spread through 5G radiation (Tuters and Knight, 2020). In combination with each other and other Covid-19 conspiracies, these theories create a toxic cocktail of misinformation (Tuters and Knight, 2020). This became clear when Twitter messages containing this conspiracy became ‘trending’ in the United
Kingdom towards the beginning of April (Ahmed and Downing, 2020). Around the same time period, over 50 phone masts in the United Kingdom were vandalized, and there were also reports on arson attacks in other parts of Europe. The most hit country on the continent was the
Netherlands, with 22 arson attacks and 3 attempted attacks (Cerulus, 2020). These attacks forced mainstream media to report on the conspiracy, which increased its platform even more.
Now, perhaps more than ever, the world is in need of clear and consistent information. On December 31st 2019, the World Health Organization (WHO) received the first message that an unknown virus was causing cases of pneumonia in Wuhan City, China. The virus spread rapidly, and on January 20 there were 278 cases of infections in China. That same day, the first cases were reported from Thailand, Japan and the Republic of Korea (World Health Organization a., 2020). With an increased death rate and an escalating number of infections worldwide, the World Health Organization quickly decided to upgrade COVID-19 to a pandemic (World Health
Organization b., 2020). In the beginning of April 2020, up to 185 countries reported cases of coronavirus infection (Aljazeera, 2020). Most countries have implemented preventative
measures, to curtail new infections. Measures such as ‘social distancing’ have a huge impact on the daily routine of citizens worldwide, and might also affect the global economy (Chen et al., 2020).
As the coronavirus spread, so did misinformation. Over the last couple of months, credible sources like the World Health Organization have been eclipsed by online hoaxes and conspiracy theories (Mian and Khan, 2020). This results in the fact that credible informational content gets drowned out by these misinformative (conspiracy) articles (Mian and Khan, 2020). As the coronavirus turned into a worldwide crisis, multiple theories have been swarming around on the internet. As mentioned earlier, one of those theories suggested a relation between the 5G network and the coronavirus. Another popular theory argued that the virus was created in a biomedical lab. Authors from Bellingcat searched for the origin of this theory, and found that it was the result of two untrustworthy, but very popular American news platforms which were shared and supported by know science sceptics (Bellingcat, 2020). In May 2020 the World Health Organization started a new campaign, together with the United Kingdom, to spread awareness on the risks of incorrect and misleading information about the coronavirus. With the campaign “Stop The Spread”, they attempt to encourage people to detect online misinformation and double check information with trusted sources, such as the World Health Organization and national health authorities (World Health Organization e., 2020). With online misinformation it is often unclear on which sources certain claims are based, and why they are shared. However, over the years, the academic interest in misinformation, disinformation and fake news has increased
(Fallis, 2009; Tandoc et al., 2018), especially since multiple presidential elections were interfered with by deliberate misinformation campaigns. This available knowledge can be used to analyze the current “infodemic” (World Health Organization e., 2020) around Covid-19, to see if comparable techniques or campaigns are used to spread misinformation online.
The spread of online misinformation can potentially harm the measures to contain the scope of the current pandemic. Experts suggest that the spread of this virus can be contained when people adjust to certain social distancing measures. In order for these measures to be
effective, people have to cooperate. Therefore it is important that the public trusts the information provided by health experts and their government. To create a better understanding of the
influence of misinformation during the coronavirus pandemic it is important to analyze its scope and the content. This thesis aims to analyze the content of online misinformation related to the coronavirus pandemic between different western countries. By comparing the type of content from different countries we might learn more about what sort of misinformative content is
currently competing with the mainstream media, and to what extent there is a difference in online misinformation between different countries. Therefore the research question that this thesis aims to answer is: “How does the content of online misinformation regarding the coronavirus differ between the Netherlands, the United Kingdom and the United States?”.
Societal relevance
Since the first warning about a new found virus in Wuhan, the news has been swarming with reports on the coronavirus. It seems like there is almost no news besides the news on the coronavirus. People are scared, because the new virus seems highly contagious and causes higher death rates than a ‘normal’ flu (World Health Organization a., 2020; World Health Organization d., 2020). The World Health Organization keeps repeating that in order to prevent and slow down the transmission of the virus, people should stay well informed about the preventative measures (World Health Organization d., 2020). However, as mentioned previously, the information on the prevention of the coronavirus gets drowned out by misinformation (Mian and Khan, 2020). In order to prevent the spread of this virus it is important that people receive scientifically approved information, that can help in effectively prevent new cases of infection. This thesis can help by analyzing the online content on the coronavirus and thereby addressing the impact of
misinformation in different countries.
Academic relevance
misinformation in relation to politics and political elections (HLEG EU Commission, 2018; Bargaoanu and Radu 2018; Guo and Vargo, 2018; Humprecht, 2019; Humprecht et al, 2020). However, the current worldwide pandemic has shown that online misinformation has become a wider problem that also affects information on other topics, such as the coronavirus (Bellingcat, 2020; Mian and Khan, 2020). This has reopened the debate on misinformation. By analyzing the content on the coronavirus and comparing it between different western countries, the current impact of online misinformation can be measured. This might provide new insight in how influential online misinformation has become. Besides, the current academic literature has been focused on misinformation in the United States. Because research based on the United States does not apply directly to European countries (Newman et al., 2017), new research on the difference between the United States and European countries should provide insight on the current state of misinformation in Europe.
Chapter 1: Literature review 1.1 Misinformation, disinformation and “fake news”
In the past years, the academic interest in online misinformation has increased (Fallis, 2009; Tandoc et al., 2018). Misinformation is information that is initially presented as truthful, that later turned out to be false (Lewandowsky et al., 2013). Another conceptualization of misinformation is incorrect information, that misinforms or misguides people (Van Kessel et al., 2020). In order to properly review previous research, it is important to establish what different terminology is used to describe misinformation and which nuances can be defined. First, authors have determined the difference between misinformation and disinformation, since these terms are used simultaneously in the literature. Fallis (2009) conceptualized the terms misinformation and disinformation by conducting a concept analysis. According to his research, the term
misinformation is mostly used when the author has made an honest mistake. On the contrary, the term disinformation is used when the author had an intent to deceive. Therefore the intentions of the source are important in determining whether or not something should be considered
misinformation or disinformation (Fallis, 2009; 2015). Misleading information can be categorized by how misleading the information is and to what extent the author intended to mislead his audience (Fallis, 2015).
In addition, both in the media and in the literature, the term fake news is frequently used to describe the phenomenon of misinformation (Tandoc et al., 2018). The term “fake news” has become a popular and commonly used word since the 2016 presidential elections in the United States. Despite its recent increase in popularity, the term fake news is not new. Fake news was
mainly used by scholars to describe satirical news commentary or tabloid journalism (Waisbord, 2018; Tandoc et al., 2018). Recently however, the meaning of the term has changed, and the term “fake news” has become an umbrella term. Despite the increase in academic interest in the topic, the definition on fake news is still ill-defined (Mourão and Robertson, 2019). Research indicated that nowadays the term fake news is used to describe six different types of misinformation: (1) news satire, (2) news parody, (3) fabrication, (4) manipulation, (5) advertising, and (6)
propaganda (Tandoc et al., 2018, p. 147). The distinction between those definitions is determined by both their level of facticity and the intent to deceive. Another important characteristic of fake news is that it is often disguised as real news. So fake news holds little facticity, has the intent to deceive its readers and is presented as legitimate news (Tandoc, 2019).
Since disinformation is shared to intentionally misinform the people who will read it, the developers need a specific motivation. Online disinformation consists of messages that are meant to serve a strategic purpose. The underlying motivation can either be ideological, commercial, or a combination of both (Tandoc et al., 2018; Humprecht, 2019). The producers of online
disinformation aim to change people’s perception on a certain topic. In the long run, the disinformation can change the opinions or decisions that the readers make (Ruokolainen and Widen, 2019). Therefore, fake news stories can be used to attract the reader's attention, to then expose them to the message that the producer wants to share (Allcott and Gentzkow, 2017). In addition, disinformation can be used to generate revenue, since polarizing or sensation news will attract attention (Tandoc et al., 2018).
When the content of disinformation was analyzed to determine the characteristics of the messages, researchers found that certain techniques were applied to the online disinformation messages. For instance, during the 2016 presidential election, most online disinformation showed moderate levels of sensationalism, clickbait, and biased, misleading content. Sensationalism is news that is intentionally evoking emotional or dramatic emotions in the beginning of the article, to grab attention for often simplified and trivialized topics (Kilgo et al., 2018 as cited in Mourão and Robertson, 2019). Clickbait is frequently used as a lure on social media. It is a technique where news headlines are designed to invoke readers to click on a link, by creating “curiosity gaps” (Blom and Hansen, 2015; Kilgo and Sinta, 2016 as cited in Mourão and Robertson, 2019). The last characteristic that is often recognized in online disinformation is strong bias. Bias in content can be recognized when the author favors one viewpoint and selects his news based on this preference (Mourão and Robertson, 2019). Sensationalism, clickbait and bias in news can create widely shared online disinformation, that is not necessarily fabricated, but very partisan and misinformative (Maurão and Robertson, 2019). Another study found that novelty is often
used to attract the people’s attention, since novel news is surprising and viewed as being more valuable, since it makes people feel like they are in on the “inside information”. (Vosoughi et al., 2018). The last technique that is often used to legitimize fabricated disinformation is repetition of the message. Pennycook and his colleagues (2018) used an experiment to determine the effect of repetition on perception of fake news. They found that: “Indeed, a single prior exposure to fake-news headlines was sufficient to measurably increase subsequent perceptions of their accuracy” (Pennycook et al., 2018, p. 1876). These findings underline the effect of the online bots, that are used to execute very effective propaganda campaigns, by swarming social media with fake news stories. By sharing the same story over and over again, they are successful in providing some semblance or legitimacy to the fake news story (Tandoc et al., 2018; Tandoc, 2019).
Another definition that is often used in the media to describe the current media
environment is: “post-truth”. The term post-truth was unknown up until roughly ten years ago, but has been used explosively for the last five years (Lewandowsky, et al., 2017). As is often the case with buzzwords, the term post-truth is utterly unclear and used in many different contexts. The term is often used to describe a world of absolute realism where the truth is relatively available. Although this claim is quite foolish, the public communication did change from information scarcity and multilayered news towards an active role for the public in the spread, production and access to news (Waisbord, 2018). Therefore perhaps another way to describe the era of post-truth, is that it marks a time where mainstream media and scientific knowledge have to compete with the “opinion market” on Twitter and other social media platforms (Anderegg et al., 2010; Lewandowksy et al., 2017).
Multiple different definitions are used to explain and describe the phenomenon of misleading information: misinformation, disinformation, fake news and post truth. Both Tandoc et al. (2018) and Fallis (2015) proposed that the difference between those definitions can be determined by the level of facticity and the intent to deceive. Misinformation is information that is false, but the author had little intent to deceive. Disinformation on the other hand is also false, but the author intentionally tried to deceive the reader (Fallis, 2015; Tandoc et al., 2018). In this thesis, the term misinformation will be used to describe the phenomenon of false or misleading information. This decision is based on the assumption that the intent to deceive cannot be determined for the articles that will be used in this analysis. Besides, the term fake news has become a buzzword in the media (Mourão and Robertson, 2019), and an umbrella term in academic literature (Tandoc et al., 2018; Mourão and Robertson, 2019), that describes multiple different types of online misinformation. Therefore the term fake news is not well defined and
somewhat confusing. Thus this thesis will focus on online misinformation in news articles on the coronavirus, and will use the term misinformation to address the issue.
1.1.1. Consequences of online misinformation
Misleading online content can have serious consequences, both societal and individual, because it can influence decisions and actions (Ruokolainen and Widen, 2019). Misinformation can make people distrust certain information or social groups (Karlova and Fisher, 2013), their government (Shin et al., 2018) and services provided by the government or governmental institutions (Lewandowsky et al., 2017). It can also affect the way people view health, science, environmental or economic related topics (Karlova and Fisher, 2013). Misinformation poses a risk, because it can undermine trust in the information society (HLEG EU Commission, 2018), and perhaps the most concerning result is a decrease in the public’s trust in science
(Lewandowsky et al., 2017). The last few years, there has been a growing disconnect between the public and science on several topics. Vaccine safety, whether or not the earth is flat and climate change have been questioned, and this has worsened in the current polarized political landscape (Mian and Khan, 2020).
1.2 Populism and misinformation
Mudde (2004) defined populism as: “an ideology that considers society to be ultimately
separated into two homogeneous and antagonistic groups, ‘the pure people’ versus ‘the corrupt elite’, and which argues that politics should be an expression of the volonté générale (general will) of the people” (Mudde, 2004, p.543). Populism has a chameleonic nature, populist politics and ideologies appear in different times and places, but they always show certain aspects
(Taggart, 2000 as cited in Mudde and Kaltwasser, 2013), that can be considered key elements of populism. There are three core concepts that define populism. First, populist ideologies are always based on the moral distinction between ‘the pure people’ and the ‘corrupt elite’ (Mudde and Kaltwasser, 2013). Although it is often unclear who ‘the pure people’ are, as it refers to an ‘imagined community’, it is often very clear who or what populists are against (Mudde, 2004). The third core aspect of populism is the appeal to the general will. Populism is often a set of ideas on how politics should work, underlining the importance of self-government (Mudde and
Kaltwasser, 2013).
Both the increase in populism and the increase in online mis- and disinformation are key threats to the functioning of our modern day democracies (Hameleers, 2020). Although these are different concepts, that have been researched separately, there are some important conceptual
similarities (e.g. Waisbord, 2018; Hameleers, 2020). First, populist communication is typically centered towards the people, conflict-focused, emotionalized and based on common sense and gut feelings (Schmuck and Hameleers, 2019; Hameleers, 2020). Although it does not mean that those types of information are necessarily false, presenting opinions and experiences of people is often less verifiable and not presented with empirical evidence. Second, populism emphasizes the central divide in politics, it’s us, ordinary people against them: the ‘corrupt’ and self-interested elites who cannot represent them. Populism is often blaming another party, elites, for the problems that ordinary people are facing. One of the elite groups that are often attacked is the media, they are accused by populist politicians of distorting reality to promote their own political agenda (Hameleers, 2020).
These findings are in line with the so called fake news effect. The psychological bias that makes conservatives describe well-known liberal media outlets such as CNN as “fake news”, and in reverse makes liberals describe well-known conservative media outlets such as Fox news as “fake news”. This development is harmful because it contributes to the so called echo-chambers, where people disregard all information that does not stroke with their moral commitments (Van der Linden, et al., 2020). Nielsen and Graves (2017) found that many people refer to fake news as “news you don’t believe”. It did not matter which media platform shared the news (Nielsen and Graves, 2017). This phenomenon is also recognized in the language of politicians, who use the term fake news as a weapon to disregard news they do not agree with (Holan, 2017). Populist politicians remain fact-free in their communication (e.g. Waisbord, 2018), and stay clear of expert knowledge and verifiable facts. Their narratives are one-sided and presented as the only reality, so opposing viewpoints by other media sources are presented as fake news (Hameleers, 2020).
Van Kessel, Sajuria and Van Hauwaert (2020) found that populist party voters’ attitudes are indeed focused on what they call people-centrism, anti-elitism and support for popular sovereignty. They fit within the image of politically disgruntled citizens. They show high levels of discontent and dissatisfaction. However, populist voters are not necessarily uninformed. So therefore, populist voters are not ignorant and they don’t show empty protest votes. Their support for a populist party might therefore be an informed and purposeful vote.
Populist groups, such as the far-right movement, have been using conspiracy theories and misinformation to promote their hateful politics (Tuters and Knight, 2020).
Bergmann (2018, as cited in Tuters and Knight, 2020) analyzed the link between populism and conspiracy theories, and he found that populist conspiracy theories often attempt to divide the world into Us versus Them. Their conspiracies try to scapegoat certain people or institutions,
while they are providing simple explanations for complex problems. It seems that the 5G coronavirus conspiracy is a very challenging one, since it managed to cater to a set of different conspiracy groups from different political spectrums. For instance, the 5G conspiracy is picked up by far-right groups, who are against technological assault by big government. On the other hand, the theory is also supported by the anti-vaccination community (Tuters and Knight, 2020). However, there is clear evidence that the spread of a lot of misinformation regarding the
coronavirus is spread by far right actors (ISD, 2020).
On April 8 the Director-General of the World Health Organization asked political leaders, to not politicize the pandemic (World Health Organization c., 2020): “For God’s sake, (...) "if you don't want more body bags, don't politicize this,”. He reacted to the harsh criticism from the United States towards the World Health Organization and China on their approach on the virus. Despite the increasing number of deaths worldwide, the virus has provoked some leaders into a blame-game (Lacina, 2020). Since there is growing concern in the United States that the
consequences of the coronavirus might strongly impact the upcoming presidential elections, President Trump has started to blame China for the current crisis (The Conversation, 2020).
1.3 Misinformation in the United States
The political polarization in the United States has increased since 1970 (Lewandowsky et al., 2017). The polarization appears to be the result of the Republican party shifting to the
political right, since the Democratic party have shifted little over the last 40 years (Hare and Poole, 2014). Polarization can be recognized in the United States’ society, since Americans now prefer to move into communities that correspond to their political preferences (Motyl et al., 2014). Besides, people also tend to choose their spouse based on their political ideology, since research shows a strong correlation, that cannot be explained by persuasion (Alford et al., 2011). The polarization of the American society also affects the way people consume and perceive news. Recent research showed that people choose their media platform based on ideology. Therefore conservatives prefer Fox News, and try to avoid CNN. Liberals and democrats showed the opposite behaviour (Iyengar and Hahn, 2009). This means that when perceiving news, liberals and conservatives inhabit two totally different worlds (Spohr, 2017).
The most dominant example of the problem with misinformation in the United States, has been the 2016 presidential election. Some suggest that President Trump wouldn’t have been elected without the influence of the misinformation stories that were going around during that time (for examples, see Parkinson 2016; Read 2016; Dewey 2016; as cited in Allcott and
news articles (Golovchenko et al., 2020). Grinberg and his colleagues (2019) found that during the presidential elections in 2016 fake news stories were shared more by, and targeted more towards conservative voters. People that shared fake news stories were people that were politically motivated, tweeted about politics themselves, and were more exposed to fake news themselves (Grinberg et al.,2019).
However, Donald Trump did not steer away from spreading misinformation himself. The independent fact check websites PolitiFact judged that 70% of all the statements Donald Trump made were false or partly false (Lewandowsky et al., 2017). President Trump has been an advocate for the fake news effect, since he started using the term as a weapon to disregard news that he did not agree with (Van der Linden et al., 2020). Since his first day in office, President Trump started tweeting about fake news about his inauguration ratings. Since then, he referred to the mainstream media as the “enemy of the people” multiple times (Van der Linden et al., 2020). Thereby increasing the polarization of the media, and the American society, and thus creating an environment in which misinformation can thrive.
1.4 Misinformation in Europe
Although the academic interest in misinformation and fake news is mainly focused on the United States, Europe didn’t escape the problem of misinformation. During the election years in France, Germany and the Netherlands there appears to be an increase in online misinformation articles. The online ‘attacks’ appear to come from Russia, and are mostly focused on pro-European parties (Scott and Eddy, 2017). On 4th February 2017 Sputnik news, a Russian propaganda website, posted a news article claiming that Emanuel Macron was a closeted homosexual. Within days the articles was referred to over a 17.000 times, on television spots, social media outlets, blogs and news articles. This incident showed how much time, energy and attention it costs to resolve and disprove such a fake claim (Hamann, 2017).
A clear example of a impactful fake news campaign is the Brexit referendum in the United Kingdom in 2016. The Brexit campaign claimed that the UK send 350 million pounds per week to the EU. This claim was very problematic, because it was completely taken out of
context. (Rose, 2017). The fake news campaign was also spread through closed Facebook groups. The accounts that spread the fake news articles were mostly connected to far-right movements, fake accounts and US or Russia oriented accounts (Spring and Webster, 2019).
So even though the focus has been on the United States, the EU is also facing challenges when it comes to misinformation. In 2018 a group of high-level experts were assigned by the European Commission, to advice on policy initiatives to counter the spread of online
misinformation (HLED EU Commission, 2018). Even though the debate on misinformation and fake news is focused on the 2016 presidential election in the United States, Europe is also facing risks and consequences. The high-level expert group listed the problems that Europe is facing, as a result of online misinformation. First, political actors can spread online misinformation
themselves to undermine the integrity of the European media systems or European institutions. There are some political leaders that show a low level of respect for media independence and therefore try to control all the different sectors of news media. This can result in skepticism amongst European citizens, with regard to their national leaders and public authorities. Second, not all news media is equally professional. While traditional news media can be of great
assistance in combating the problem of misinformation, some news media contribute to the problem. This weakens the overall trust in media amongst Europeans. Third, an important factor in the spread of online misinformation is individuals. Citizens individually or collectively share false information or ideologically motivated misinformation. Fourth, the large digital social media platforms add to the scale and news ways that misinformation can be spread nowadays (HLED EU Commission, 2018). These problems underline the complexity of dealing with online misinformation. Misinformation presents itself in many forms and can harm European values, politics and trust.
Chapter 2: Theoretical framework 2.1 Defining misinformation
In order to analyze the different types of misinformation it is important to determine the exact definitions of these concepts. Fist, misinformation will be understood as information that is initially presented as truthful, that later turned out to be false (Lewandowsky et al., 2013). To analyze different types of online misinformation, the conceptualization that is provided by Tandoc will be used (Tandoc et al., 2018; Tandoc, 2019). By reviewing published articles that used the term fake news to describe online misinformation, Tandoc and his colleagues found that nowadays the term fake news is used to describe six different types of misinformation: (1) news satire, (2) news parody, (3) fabrication, (4) manipulation, (5) advertising, and (6) propaganda. They used two different dimensions to categorize these different types of misinformation: ‘level
of facticity’ and ‘author’s immediate intention to deceive’. The categorization is shown in table 1.
This table is directly taken from their article.
The first type of misinformation, that Tandoc and his colleagues identified is news satire. This type of misinformation actually was the most commonly used operationalization. News satire refers to mocking news programs, that use humor to reflect on the news. An example of
these type of shows is The Daily Show in the United States (Baym, 2005 as cited in Tandoc et al., 2018). They use the setup of a regular news program, so it’s often a “talking head” behind a desk. They reflect on real news, but use satirical humor to entertain their audience. Therefore their level of facticity is high, however their intent is to entertain, rather than deceive (Tandoc et al., 2018). The second type of misinformation is news parody. News parody also relies on humor to draw in their audience. They also use a format that is very similar to mainstream news media. The difference with news satire is that news parody injects ridiculous, non-fictional information to mock current affairs. Instead of reflecting on the news, they create entirely fictitious news stories. A clear example of news parody is the American website The Onion. Although they do get mistaken with real media, their level of facticity is low, as well as their intent to deceive their audience (Tandoc et al., 2018).
Third, the authors identified news fabrication as an operationalization of misinformation. This refers to articles that are published in the style of regular news, but with no actual factual basis. The difference with news parody is that there is no understanding between the publisher and the reader that the content is nonfactual. Therefore, the intention with news fabrication is often to misinform the reader. News fabrication is using the technique of republishing to feign legitimacy. Their level of facticity is low, and their intent to deceive high (Tandoc et al., 2018). The fourth operationalization of misinformation is photo manipulation. With this type of fake news, photos or videos are used to create a false narrative. Unfortunately, the manipulation of photos has become very common, since the digitalization of photography and the improvement of photoshop technology. Manipulation consists of both the tampering with images to create a different narrative and the use of pictures in different contexts to change the narrative. Therefore, the picture might be factual, but by changing its context, the intent to deceive is high (Tandoc et al., 2018).
The fifth type of misinformation is advertising and public relations. This
operationalization describes advertisements and press releases, that are disguised as genuine news reports. Public relation firms use pre-packages video segments to promote their product or
ideology, by inserting them into news media (Farsetta and Price 2006, 5, as cited in Tandoc et al., 2018). Another example is native advertising, a format which serves both the intent to inform and to advertise. This type of advertisement is focusing on facts, however, the facts are used to
advertise the product and are therefore often incomplete (Deziel, 2014 as cited in Tandoc et al., 2018). The last form of advertising is what is currently known as “clickbait”. This practice uses remarkable titles and images to persuade people to click on the item (Tandoc et al., 2018). The last type of misinformation is propaganda, a type of misinformation that has often
been researched over the last years. Propaganda is news that is created by political entities to influence public perceptions of certain topics. Although there is an overlap between
advertisement and propaganda, the difference can be found in the motivation. While propaganda has the intent to persuade people, the intent with advertising is mainly financial (Tandoc et al., 2018).
Table 1
A typology of fake news definitions
Author’s immediate intention to deceive
Level of facticity High Low
High Native advertising
Propaganda Manipulation
News satire
Low Fabrication News parody
Retrieved from: Tandoc Jr, E. C., Lim, Z. W., & Ling, R. (2018). Defining “fake news” A typology of scholarly definitions. Digital journalism, 6(2), 137-153.
For this study the six different types of misinformation will be used to analyze the differences in fake news content between the Netherlands, the United Kingdom and the United States. Since the United States has had the most problems with misinformation in the past, especially during the 2016 presidential elections (Tandoc et al., 2018; Lewandowsky et al., 2017), the expectation is that in United States there are more different types of misinformation than in the United Kingdom and the Netherlands.
2.2 Defining resilience
This thesis tries to compare the nature and quantity of online misinformation regarding the coronavirus. In order to be able to compare different countries, it is important to use a suitable theoretical framework.
Humprecht, Esser and Van Aelst (2020) came up with a framework for cross-national comparative research. Based on a literature review, they were able to identify seven indicators that influence a country’s resilience to online misinformation. First of all, their definition of resilience is based on a structural context in which misinformation does not reach a large group of people, or in which people are less inclined to share and support misinformation. As
populist communication, (3) low trust in news, (4) weak public service media, (5) more fragmented, les overlapping audiences, (6) large ad market size and (7) high social media use. (Humprecht et al., 2020).
The first indicator is based on multiple studies that argued that polarization is an important driver for the spread of online misinformation (e.g. Allcott and
Gentzkow, 2017). Polarization can be understand as the separation of politicians and elites on issues or political spectrums (Dalton, 2008; Hetherington, 2001 as cited in Humprecht et al., 2020). Strong partisanship provides an environment in which people only share and consume news that is favorable of their political preference (Muddiman and Stroud, 2017). Because there is a divide in information available, it becomes very difficult for people to determine whether or not news is legitimate or not. Thus the assumption is made that societal polarization decreases the resilience to online misinformation (Humprecht et al., 2020).
The second indicator is populist communication. As mentioned previously in the current study, scholars have found a link between populism and misinformation, since they share characteristic psychological underpinnings (e.g. Waisbord, 2018; Hameleers, 2020). Populist communication often uses partisan language and misinformation to strengthen their Us versus Them narrative. Thus countries with high levels of populist communication are less resilient to online misinformation (Humpecht et al., 2020).
The third indicator of resilience is low trust in media, and thus refers to the media environment in countries. Low trust in mainstream media has a crucial impact on how people perceive information and how they interpret certain problems (Curran et al., 2012; Van Aelst et al., 2017). It also increases the use of alternative sources for news consumption, such as social media (Tsfati and Cappella, 2003). Low trust in media therefore lowers the resilience to online misinformation (Humprecht et al., 2020).
The fourth indicator is weak public service media, which refers to the availability of public service broadcasting (PBS). Research shows that countries with PSB, have high levels of ‘hard news’, and therefore their citizens are better informed. Environments with weak public broadcasting are therefore assumed to be less resilient to online misinformation Humprecht et al., 2020).
More fragmented, less overlapping audiences are identified as the fifth indicator of resilience. This indicator is similar to the first indicator, since countries with a strong social polarization tend to also show a divide in news consumption. Humprecht (et al., 2020) assumes that large overlap in news consumption reduces the risk of the confrontation with misinformation. The sixth indicator is a large add market size. Creating and spreading misinformation can
be very lucrative, due to add revenue. During the presidential elections in the United States in 2016, Macedonian “fake news factories” were produced to pursuit add revenue (Nielsen and Graves 2017; Humprecht et al., 2020). Since misinformation often contains sensationalist and emotionalized topics and stories, it is created to attract attention and clicks. Thus the marked size influences the resilience to online misinformation (Humprecht et al., 2020).
The seventh indicator is identified as high social media use, since social media is considered an loudspeaker for online misinformation (Meraz and Papacharissi,
2013; Shin et al., 2017; Singer, 2014 as cited in Humprecht et al., 2020). In countries with many social media users, it is easier to distribute false information or manipulated images. Therefore countries with a lot of social media users are perceived as less resilient to online misinformation (Humprecht et al., 2020).
By analyzing data from these countries, and by applying the seven indicators, the authors were able to establish three different clusters of countries, based on their resilience to online misinformation. Cluster one consists of countries that showed high resilience to online misinformation. Countries in this cluster are Northern and Western European countries, plus Canada. Cluster two consists of Greece, Italy, Portugal and Spain. These countries were all characterized by polarized-pluralist media systems. The third cluster consists of just the United States, as they show an exceptional role in the context of online misinformation (Humprecht et al., 2020). Based on the assumptions that there are seven different types of online misinformation (Tandoc et al., 2018), and that countries are different in their resilience to online misinformation (Humprecht et al., 2020), this thesis will address the following two hypotheses:
H1: Countries with low resilience to online misinformation will show more different types of misinformation when compared to countries with high resilience.
H2: Countries with low resilience to online misinformation will have more politically motivated sources that share misinformation than countries with high resilience.
Chapter 3: Method
The research question and hypothesis in this thesis will be answered by conducting a qualitative content analysis. This research design will allow the comparison of messages about misinformation and the coronavirus published by news media in different western countries. The analysis is based on the method used in the content analysis conducted by Humprecht (2019). She compared four different western countries, that recently held political elections, on the online misinformation content. The method proved sufficient to compare cross-national differences in
online misinformation, and will therefore be used to conduct the analysis for this research. The steps that were taken to determine what content will be included in this analysis will be explained in this method section.
3.1 Study setting and sample selection
The first step was to determine which countries would be selected for this comparison. Three countries were selected, based on their level of resilience to online misinformation. The countries were selected from the comparative research by Humprecht, Esser and Van Aelst (2020), since they provided a clear framework for cross-national comparative research on online misinformation. Because both hypotheses are based on the connection between the resilience to online misinformation and the content of online misinformation, we will include countries that score differently on the framework provided by Humprecht, Esser and Van Aelst (2020). First, the Netherlands was selected, based on their high resilience to online misinformation. Second, the United Kingdom was selected. Even though this country was also paired in the cluster containing countries with high resilience, the United Kingdom showed major misinformation problems during the Brexit campaign (Humprecht et al., 2020). Because the Brexit referendum increased the polarization of the political landscape in the United Kingdom, this country will also be relevant to compare. The third selected country is the one country from the third cluster, the United States. The decision to not include a country from the second, southern European cluster is based on linguistic limitations.
The second step was to select two nationwide news media outlet for each of the included countries. Even though the three included countries have very different media environments, the attempt was made to include comparable or equivalent media outlets. The media outlets were selected based on their nationwide coverage and readership, and on their availability of online content. Both the United Kingdom and the Netherlands have a public broadcast corporation, that is funded by the state. Therefore the BBC (UK) and the NOS (NL) are both included in the sample. Because the United States don’t have a public broadcasting corporation, the most international broadcasting corporation, CNN, was included in the sample. Besides, because in this research the focus will be on online misinformation, the media outlets that were selected had to report on online misinformation stories or be active in fact checking certain stories with regard to the coronavirus. Preferably, we would have been able to select fact checking websites for all of the three countries, however, in the Netherlands there are no active fact checking websites, or at least none that reported enough articles to be included in the research. Therefore the decision was made to include media platforms that are not exclusively fact checking websites, but that are
engaging in fact checking news. Even though all the included websites are actively fact checking, the most active fact checking outlets for every country were selected for the sample. The websites that were selected for the three different countries are presented in table 2.
The third step was to select relevant news articles from the fact checking websites. The articles that were selected had to be published between December 31 2019 and may 1st 2020. The dates are based on the first day the World Health Organization was informed about the virus (World Health Organization a., 2020). If subpages on fact checks were available, the subpages were used to select articles on the topic. When there were no subpages available the search term ‘fact check’ or the Dutch equivalent ‘nepnieuws’ was used to select the news stories. The next step was to select the news articles with misinformation on the coronavirus. Some websites provided subpages with fact checks regarding the coronavirus, but when these pages were not available the articles were selected based on their titles. The titles were scanned for the terms corona, coronavirus or covid-19. In order to randomly select news stories, every other article was selected from the articles that were presented after the search, unless there weren’t enough articles to apply this strategy. Based on the availability of articles, and the capacity of the research, up to 10 articles were selected from each website. The summary table with all the included articles is added in the appendix. The included media platforms and the number of included articles is presented in table 2.
Table 2
Selection of fact checking websites
Country Name of website Number of stories selected
The Netherlands - Nu.nl - NOS
10 6 The United Kingdom - The Guardian
- BBC
10 10 The United States - Washington Post
- CNN
10 10
Based on: Humprecht, E. (2019). Where ‘fake news’ flourishes: a comparison across four Western democracies. Information, Communication & Society, 22(13), 1973-1988.
3.2 Measurements
Although this research will be conducted by a quantitative content analysis, the first step will be to select a subsample of articles to conduct a small qualitative content analysis. The subsample will be analyzed to identify the dominant topics, that will be used for the qualitative
part of the content analysis. As mentioned earlier, all articles on the coronavirus will be included, so therefore it is important to determine which dominant topics are presented. Most of the time, articles present the dominant topic in the title, so therefore to determine the dominant topics, the titles of all the articles from the subsample were scanned.
The categories for the quantitative part of the analysis were again based on the categories presented by Humprecht (2019). So the articles will be coded on what type of misinformation they present (Tandoc et al., 2018), most important topics, sources and the targets. These categories are relevant and will eventually help with answering the research question and hypotheses that were presented.
First, the articles will be placed within one of the six different types of online
misinformation. In this analysis, all six different types will be included, since all these different types of misinformation can be reported on by fact check sites. Besides, it is unclear what type of misinformation is dominant during the coronavirus pandemic, therefore all six types will be included. So they will all be coded in one of the following types: (1) news satire, (2) news parody, (3) fabrication, (4) manipulation, (5) advertising, and (6) propaganda (Tandoc et al., 2017).
Second, the dominant topics will be coded, based on the qualitative research. This step is important, because it will help with answering the research question. The dominant topic can be retrieved from the title of the article, key-words or in the subtitle. The following dominant topics were selected: Politics (e.g. measures), Health (e.g. remedies), Vaccination, 5G network,
Economy, Origin of the virus and Others.
Third, the sources of each news article will be coded. For this research the source will be the one who is introducing or sharing the information, not the source that is mentioned in the article. This decision is made on the assumption that the sources mentioned in the article are not always legitimate. The sources will also be categorized, to assure that they can be compared. Therefore the sources will be categorized as: politicians/political institutions, national
government (e.g. the president, state reports), business/interest groups (e.g. InfoWars, anti-Vax), journalist/blogger, social media (e.g. influencers, Facebook groups, WhatsApp) and other (e.g. rumors, anonymous).
The final step will be to code the articles based on their target objects. News stories are mostly targeted towards a topic, person or institution (Humprecht, 2019). In order to code the articles with online misinformation on the coronavirus, the following categories were included in the codebook: politicians/political institutions (e.g. the president, political parties), national government (e.g. government officials, measures imposed by the government), China/Chinese
government, America/American government, Health/Lifestyle (e.g. health decisions by society), pharmaceutical industry, 5G (e.g. network providers, radiation), individuals (e.g. Bill Gates), others.
Chapter 4: Analysis 4.1 Sample description
For this content analysis, 56 news stories were selected from the fact checking sections of news media websites. The articles were selected between 14 may 2020 and 25 June 2020. The aim was to include 10 articles from every news website. Unfortunately there were only 6 articles on misinformation published on the NOS website, between 31 January 2020 and 1 may 2020 (NL). Therefore the number of included news articles from the Netherlands is lower than those from the United Kingdom and the United States. Although the sample selected from the United Kingdom did also include two articles that were included based on their title, that didn’t contain any specific information on misinformation. In total there were 162 different fact checked news stories coded from the included articles, since some articles reported on multiple stories. The Dutch news media reported a total of 45 stories containing misinformation, the United Kingdom reported on a total of 46 different stories and the United States news media reported on 71 news stories that contained misinformation.
4.2 Cross-national analysis
4.2.1 Sources of online misinformation
The results of the cross-national content analysis show clear differences between the content of online misinformation regarding the coronavirus. The sources of online
misinformation show a lot of differences between the three different countries. The websites that were analyzed from the United States show that the largest share of stories that were fact
checked, were stories that came from their national government (59% of all the stories from the US). The websites from the United Kingdom did also report on a large share of stories that came from a government source (26%). However, both the stories from the US and the UK that were included in this category mostly reported on false statements from President Trump or American government officials. In contrast, news websites from the Netherlands reported no stories that were directly shared by their national government, or a government related author. In the Netherlands there was very little reporting on Donald Trump.
Table 3
Sources of online misinformation
The Netherlands The United Kingdom The United States Politicians/Political institutions 1 (2%) 1 (2%) 4 (6%) National government 0 12 (26%) 59 (83%) Business/Interest groups 5 (11%) 3 (7%) 0 Journalist/Blogger 7 (16%) 2 (4%) 0 Social media 31 (69%) 25 (54%) 3 (4%) Other 1 (2%) 3 (7%) 5 (7%)
Websites from The Netherlands and the United Kingdom published mostly about stories that came from social media (69% of all the stories in the Netherlands, and 25% from the UK). In contrast, in the US only 4% of all stories originated from social media sources. These stories contained a lot of rumors, conspiracies and speculation from social media accounts from social media platforms such as Twitter, Facebook and Instagram. Another frequently reported type of social media source is shared group messages through WhatsApp. In the United States there were five reports (7%) on stories from sources that couldn’t be included in any of the categories. These stories contained widespread gossip stories, that were mentioned in the fact check articles,
without any identified sources.
4.2.2 Topics of online misinformation
The differences in the most reported topics is interesting and this strongly suggests a difference in resilience between the three included countries. In the United States, the stories were mainly focused on politics (48% of all US stories). The US stories were often fact checking statements from the President or his government officials on topics such as the lock down
measures, testing capacity and flight restrictions. In the United Kingdom there were also stories on politics (17%), but again these articles were mainly reporting on American politics. In The Netherlands there we no articles included that reported on politics. In The Netherlands, and the same counts for the United Kingdom, most misinformation was focused on health related topics (44% of all NL stories, 39% of all UK stories). The articles on health were also a dominant topic in the United States, since 35% of all the stories were on health. The articles that were included in
this category mainly contained health advices, suggested remedies and theories on the seriousness and infectiousness of the coronavirus. It seems that there are a lot of different stories containing false or harmful health advice circulation on social media. In all the included countries the fact checkers were occupied with setting these stories straight. The three most popular conspiracy topics are on the origin of the virus, the dangers of a vaccine and the influence of the 5G network on the virus. In the Netherlands, a large share of stories was reporting on these conspiracy topics, 27% of all the stories were about the origin of the virus, 13% on vaccinations and 11% on the 5G network.
Table 4
Topics of online misinformation
The Netherlands The United Kingdom The United States
Politics 0 8 (17%) 34 (48%)
Health 20 (44%) 18 (39%) 25 (35%)
Vaccination 6 (13%) 3 (7%) 1 (1%)
5G network 5 (11%) 7 (15%) 0
Economy 0 0 3 (4%)
Origin of the virus 12 (27%) 4 (9%) 4 (6%)
Other 2 (4%) 6 (13%) 4 (6%)
In the United Kingdom, these conspiracy topics contained a smaller number of articles, 15% on the 5G network, 9% on the origin of the virus and 7% on vaccinations. In the US there was only a small number of stories on these topics (6% on the origin, 1% vaccination, non on 5G). The stories that were included in the category Other were for instance on a message that was spread via WhatsApp in the United Kingdom, telling people to shut their doors and windows, because helicopters would come and spray disinfectant on their city (Goodman, 4 April 2020).
4.2.3 Targets of online misinformation
Across all countries most of the online misinformation targets health and lifestyle decisions that people make or need to make. These health advices are often proposing clear and easy solutions to cure Covid-19 or prevent people from getting infected with the coronavirus (NL 40%, UK 30%, US 25%). However, in the United States, most articles were targeting their national government (30% of all the US articles). It is worth noting however, that most of these
contained accusations of President Trump, targeted towards his own administration and their imposed measures. The category other is also containing a large share of the United States’ articles (19% of all articles). The large proportion of stories from the United States that are coded as Other can be explained by the nature of a lot of false statements by President Trump, that were reported as fake news by the fact check sites.
Table 5
Targets of online misinformation
The Netherlands The United Kingdom The United States Politicians/Political institutions 1 (2%) 3 (7%) 8 (11%) National government 4 (9%) 5 (11%) 21 (30%) China/Chinese government 4 (9%) 5 (11%) 5 (7%) America/American government 1 (2%) 2 (4%) - Health/Lifestyle 18 (40%) 14 (30%) 18 (25%) Pharmaceutical industry 2 (4%) 2 (4%) 0 Individuals 8 (18%) 3 (7%) 0 5G 5 (11%) 6 (13%) 0 Other 2 (4%) 6 (13%) 19 (27%)
Most of president Donald Trumps’ statements are not directly targeted towards anything or anyone, they are however often false and misleading. It is also worth mentioning that the stories that were included in the category individuals, were mostly targeting Bill Gates. Gates was often accused of developing the virus to suppress the world's population. Besides, he would also be developing a vaccine with a malicious microchip inside to track all citizens by using the 5G network. Across all three countries, online misinformation on the coronavirus focuses on health related topics and is targeted towards health and lifestyle choices the people should make. Therefore targeting citizens, measures and cures.
4.2.4 Type of online misinformation
Finally, to answer the first hypothesis the articles were coded on what type of
misinformation they were reporting on. The countries show no significant differences in types of online misinformation, however the United States show a lot more propaganda stories (51%), than the other two countries (NL 2%, UK 20%). However, this difference can be explained by the fact that the United States’ websites mostly reported on statements from President Trump. Since the online misinformation articles from the Netherlands and the United Kingdom were mostly about fake health advice, most of their stories contained fabricated misinformation (NL 91%, UK 54%). However, since there is no significant difference between the three countries, the
hypothesis cannot be confirmed.
Table 6
Type of online misinformation
The Netherlands The United Kingdom The United States
News Satire 0 0 0 News Parody 0 0 0 Fabrication 41 (91%) 25 (54%) 20 (28%) Manipulation 3 (7%) 8 (17%) 0 Advertising 0 4 (9%) 0 Propaganda 1 (2%) 9 (20%) 51 (72%)
4.2.5 Type of online misinformation and source of online misinformation
As is previously explained, the results from this content analysis show large differences in the spread of online misinformation between the three different countries. Because the results showed that both fabrication and propaganda are frequently reported as the type of
misinformation, and that social media and the national government are frequently reported as the source, it is interesting to compare these two categories. The results are presented in table 7. As the results show, of all the articles that were shared through social media, the majority consisted of fabricated online misinformation, 50 of the 58 articles. The other way around, of all the articles that consisted of fabricated news, 58% was shared through social media. 15% of all the fabricated stories came from the national government, which in this case meant the American government. Of all the propaganda articles, 93% was shared through the national government, again this meant the American national government. The misinformation consisted mainly out of
only four articles containing misinformation that is categorized as advertising, it is notable that 75% of those were spread by interest groups. An example of an interest group, is InfoWars, the online conspiracy news website/podcast, hosted by Alex Jones. Jones spread multiple stories that consisted of misinformation with the intent to sell his own miracle cures. He for instance tried to advertise his silver products as cures for Covid-19, even though the American Food and Drug Administration (FDA) had warned him multiple times to stop selling these products (Haasch, 2020). The articles that reported on manipulated misinformation were mainly shared through social media (64%), but as with the advertising articles, there were very little articles reported containing this type of misinformation.
Table 7
Type of online misinformation x source of online misinformation
Fabrication Manipulation Advertising Propaganda Total
Politicians 3 (4%) 1 (9%) 0 3 (5%) 7 National government 13 (15%) 1 (9%) 0 57 (93%) 71 Interest groups 4 (5%) 0 3 (75%) 1 (2%) 8 Journalist 7 (8%) 2 (18%) 0 0 9 Social Media 50 (58%) 7 (64%) 1 (25%) 0 58 Other 9 (11%) 0 0 0 9 Total 86 11 4 61 162 Chapter 5: Discussion
The issue of online misinformation seems to be a worldwide problem, however, some countries have had more problems with misinformation than others. This study aims to analyze the differences in online misinformation on the coronavirus between different countries. Therefore the content of online misinformation about the coronavirus was analyzed and
compared between the Netherlands, the United Kingdom and the United States. The aim of this analysis was to answer the proposed research question that stated: “How does the content of online misinformation regarding the coronavirus differ between the Netherlands, the United Kingdom and the United States?”. Before the research question will be addressed, the two hypotheses will be addressed in order.
Based on the literature, the assumption was made that countries with low resilience to online misinformation will show more types of misinformation when compared to countries with high resilience (H1). This assumption was based on the research by Humprecht et al. (2020), who found that countries with high resilience to online misinformation are well equipped to face misinformation. Those countries have trusted media outlets, that enable citizens to obtain their information there and uncover misinformation. Therefore, certain types of misinformation should be filtered out, since people are well equipped to recognize fake articles. However, the results paint a different picture than expected. In contrast to the hypothesis, the included articles from the United States only reported on two different types of online misinformation. 72% of the
misinformation contained propaganda, the remaining 28% of the articles contained fabricated news. Therefore the first hypothesis, the assumption that countries with low resilience to online misinformation will report more types of misinformation, cannot be confirmed by this research. However, the large coverage of propaganda in the United States is in line with previous research that suggests that polarization leads to more misinformation from political entities (Tandoc et al., 2018). In combination with populist communication this can create an environment where online misinformation can flourish (Humprecht et al., 2020; Lewandowsky et al., 2017; Van der Linden et al., 2020). The underreporting of other types of misinformation might be the result of
overshadowing. Perhaps because the United States are currently in an election year, or because the media is set on correcting the President, but it is possible that other fake news stories are simply overshadowed by news surrounding politics. It is unlikely that the results of the current study suggest that these other types of online misinformation are not present in the United States. It is, therefore, more likely that another concept explains the lack of different types of
misinformation.
The second hypothesis was based on the assumption that countries with low resilience to online misinformation will have more politically motivated sources that share misinformation than countries with high resilience. In this analysis, the country with the lowest resilience was the United States. As presented in table 3, 83% of all the misinformation articles that were shared in the United States, came from their national government. 7% came from politicians or political institutions. In contrast, in the United Kingdom only 26% of all the articles came from the national government, and only 2% from politicians. In the Netherlands, the country with the highest resilience to online misinformation, there were no reports on articles that came from the national government, and only 2% came from politicians. The differences in sources from online misinformation between the three included countries seem to confirm the second hypothesis. The results seem to be in line with the literature on this topic. The political environment in
