Abstract
This study explores how fact-checkers understand information disorder in Ibero-America, in particular the COVID-19 disinformation. We conducted a quantitative content analysis of the LatamChequea Coronavirus alliance database and in-depth interviews with journalists from the network. Evidence found that one of the most prevalent disinformation topics was the government’s restrictive measures, threatening to jeopardize the effectiveness of public health campaigns. This, added to disinformation that eroded the trust in the institutions and the press, and the opacity of governments constituted a political crisis in Ibero-America. Under this scenario, fact-checkers created relevant journalistic collaborations and strategies to fight disinformation in the region.
Introduction
The COVID-19 outbreak unleashed a social, economic, political, and public health crisis. A crisis is an episode, an event, or a complex process that, if not resolved, can complicate the development of a system and/or the role of a group or an individual within it (McNair, 2017). As a political crisis, the pandemic involved a series of political actions to contain the virus. These included lockdowns, border controls, quarantines, and information on the virus (Dodds et al., 2020).
The director of the World Health Organization (WHO) in 2020, Adhanom Ghebreyesus, equated the danger of COVID-19 with the impact of spreading false information about the disease, its effects, and its treatment, declaring that “we’re not just fighting an epidemic; we’re fighting an infodemic.” The WHO had already observed that the outbreaks of Zika virus, Ebola, and cholera were accompanied by an infodemic, a “damaging ‘epidemic’ of rumours” (World Health Assembly, 2019). The danger of viral disinformation is its influence on the effectiveness of public health measures taken by governments in crisis contexts (Papapicco, 2020).
This crisis scenario led to not only the dissemination of disinformation, but also its politicization (Gehrke & Benetti, 2021). Politicization is to be understood here as the transfer of an issue to the political sphere (Zürn, 2019). In this politicized crisis context, fact-checking journalism emerges as an effective tool to fight the infodemic.
This article seeks to reveal how fact-checkers understand information disorder in Ibero-America, in particular disinformation about COVID-19, and what strategies they employ to address it in crisis contexts. To understand this information disorder, we first identified the prevalent interpretations and the characteristics of the COVID-19-related disinformation (i.e., topics, formats, sources, etc.) through a quantitative content analysis of the LatamChequea Coronavirus alliance database. Then, we performed in-depth interviews with the journalists belonging to the network to comprehend how the characteristics of disinformation were related to the social and political context of each country, and how collaborative alliances could tackle the spread of disinformation in the region. This article aims to fill a gap in journalistic practices of fact-checking in Ibero-America in periods of crisis.
Information Disorder
The new information environment is the result of the digitalization of journalistic production and routines, the widespread use of the internet and social media, the rapid expansion of mobile technology, and the world economic crisis. Without hesitation, “the more sudden change in the information landscape is the rapidly expanding abilities of a wide array of actors to sow doubt through misinformation, circulate conspiracy theories, provide ‘alternative facts,’ or otherwise question the accounts of news media” (Ekström et al., 2020, p. 206). All of the above are causes of the three types of information disorder: misinformation (false information that is shared with no intention of causing harm), disinformation (false information that is shared to harm), and malinformation (truthful information, which should not have left the private sphere that is shared to harm) (Dan, 2021; Wardle & Derakhshan, 2017). Disinformation as a type of information disorder generates an atmosphere of mistrust toward institutions (Rubin, 2019), politicians, and the media (McNair, 2017; Nielsen & Graves, 2017).
Trust in journalism is based on the reliability, credibility, and responsiveness precepts (Brants, 2013). The absence of some of these precepts generates skepticism toward the media, “a subjective feeling of mistrust” (Tsfati & Cappella, 2003, p. 506). This trust erosion could explain why citizens view the media as a powerful institution aligned with governments (Brants, 2013).
In some Ibero-American countries, journalists have tried to overcome that credibility crisis and the precariousness of the journalistic practice by promoting mostly digital media, which Manfredi Sánchez et al. (2015) call “entrepreneurial journalism.” Fact-checking organizations are an example of these types of digital media. They are generally small companies that, to survive, seek new business models and forms of cooperation with other media to optimize resources and have greater journalistic coverage.
Fact-Checking Journalism
Fact-checking is described as “an evaluation of a claim or a text that is already public, rather than something one intends to make public” (Amazeen, 2019, p. 4). Despite the complexity of establishing the origins of fact-checking in journalism and how the process of verifying facts was carried out before the internet (Fabry, 2017), it is worth noting that fact-checking was an intrinsic process in information media, always related to the ideal of journalistic impartiality. In the past, journalists whose job was to scrutinize articles for accuracy before publication were not called fact-checkers, but researchers, focusing on sources and data verification in public libraries.
The origins of modern fact-checking go back to the early 20th century in the United States (Amazeen, 2020) being configured as a journalism genre (Graves, 2016) when, in the wake of the internet and the convergence of digital technologies, digital journalism was becoming popular around the world.
In 2015, before the elections in the United States that put Donald Trump in the White House, the Duke Reporters Lab journalism research center listed 64 fact-checking organizations around the world. At the beginning of 2020, the number had increased to 237 associations in 78 countries engaged in verifying the claims of public figures, flagging false information, and explaining current events.
Fact-checking has had a rapid expansion on a global level (Vu et al., 2023). According to Amazeen (2019), the reasons behind this expansion are the decline in journalism practices, which went from a public service to a business venture, public disempowerment, technological changes, and social/political crisis and reform movements.
Also in 2015, the Poynter Institute, a professional journalism school in Florida, set up the International Fact-Checking Network (IFCN) to assemble the developing community of fact-checking organizations worldwide. By the end of April 2020, 71 organizations were carrying the IFCN badge, a signatory status certifying that they embrace the commitment to five principles: non-partisanship and fairness, transparency of sources, transparency of funding and organization, transparency of methodology, and the commitment to an open and honest corrections policy.
The IFCN promotes constant training and skills development within the network. Fact-checkers also receive skills training and resources from social media platforms. They are regularly adapting their verification protocols to the fast-changing media landscape and the regulation and uses of social networks since disinformation circulates through social media faster than verified news (Vosoughi et al., 2018).
Fact-Checking Organizations in a United Front
Networks like the IFCN that bring together transnational fact-checking organizations (Higgins Joyce & Harlow, 2020) are based on collaborative journalism (Hellmueller et al., 2019; Stonbely, 2017). Collaborative journalism is defined as a “cooperative arrangement (formal or informal)” (Stonbely, 2017, p. 14) among media associations that intend to amplify and empower the work that each organization performs. The collaborations can occur at local, national, or international levels. Collaborative practices between fact-checking organizations promote accountability in journalism. “Fact-checkers can materialize partnerships that help to reduce production costs, share content across platforms and facilitate the dissemination of verifications in different countries” (Moreno-Gil et al., 2021, p. 260).
Collaboration is an intrinsic characteristic of journalism just like competition. Competition has always had greater visibility due to the business logic that has governed the journalistic field (Graves & Konieczna, 2015). However, collaboration has changed the common mind-set of rivalry and allowed fact-checking organizations to demonstrate the growing advantages of cooperation (Vu et al., 2023).
A particular cross-border collaboration is the LatamChequea network, which gathers Spanish- and Portuguese-speaking fact-checkers in South America, Central America, Mexico, Spain, and Portugal. This network of Ibero-American fact-checkers shares cultural backgrounds, languages, and borders. Most of these organizations were inspired and supported by Chequeado, a digital media pioneer in data verification in Latin America. Chequeado was founded in Argentina in 2010. Since then, they have been training and exporting their analysis methodology elsewhere.
Despite the fact that “the practice of national and transnational collaboration in the region is still scarce” (Cueva Chacón & Saldaña, 2020, p. 14), the pandemic motivated the creation of the LatamChequea Coronavirus alliance to specifically flag COVID-19 disinformation. The project, which emerged from the LatamChequea network initiative, links 35 organizations in 18 countries. The alliance intended to accelerate the process of verification because “in many cases, the disinformation that circulates is the same in different countries, and being able to count on the work of others helps to debunk falsehoods faster and prevent their spread” (LatamChequea-Coronavirus, 2020).
Since the beginning of the pandemic, the LatamChequea Coronavirus network created a collaborative database where the organization’s members registered the checks (in Spanish and Portuguese) performed in each country of the region. The LatamChequea Coronavirus database complemented the global CoronaVirusFact Alliance database, a joint database nurtured by over 100 fact-checking organizations coordinated by the IFCN.
The LatamChequea Coronavirus database has been used in several quantitative studies (Gehrke & Benetti, 2021; López-Pujalte & Nuño-Moral, 2020; Salaverría et al., 2020; Sánchez-Duarte & Magallón Rosa, 2020). For instance, Salaverría et al. (2020), and Sánchez-Duarte and Magallón Rosa (2020) studied the hoaxes debunked by Spanish fact-checking organizations. Gehrke and Benetti (2021) analyzed the disinformation that circulated in Brazil. López-Pujalte and Nuño-Moral (2020) examined the disinformation, the explainers, the public figures’ discourses, and the media claims in Ibero-America. None of them aimed to study the characteristics of collaborative work, nor the shared strategies that fact-checkers implemented to halt disinformation.
Similar to LatamChequea, other transnational fact-checking initiatives exist. The report on Freedom of Expression and Addressing Disinformation on the internet (Bontcheva & Posetti, 2020) categorizes the cross-border fact-checking responses at the global and regional level. On a global level, apart from the IFCN and the Duke University Reporter’s Lab Database, the authors also mention First Draft, a transborder independent network of journalists.
Besides the traditional English-speaking fact-checking projects that have dominated the scene, other regional alliances have emerged. These include AfricaCheck, a key agent in Africa with a presence in South Africa, Kenya, Nigeria, and Senegal; the Arab Fact-Checkers Network, which was born as a response to the overload of COVID-19-disinformation in the Arab region; the Social Observatory for Disinformation and Social Media Analysis (SOMA) founded by the European Commission to support fact-checkers, news media, academics, and activists in Europe; and Full Fact, an independent team of fact-checkers in the United Kingdom.
This article was guided by the following general research questions:
Methods and Sample
A mixed methodology combining quantitative and qualitative techniques was used for this research, which had two phases. First, we conducted a secondary quantitative content analysis on the LatamChequea Coronavirus database to determine the characteristics of the COVID-19 fact-checked content. The second phase entailed a primary qualitative analysis through in-depth interviews with Ibero-American fact-checkers from the LatamChequea Coronavirus network. The interviews aimed to explore how fact-checkers understand information disorder in Ibero-America.
First Phase: Secondary Analysis
The secondary analysis follows the structure proposed by Radey (2019), namely: (a) selecting a data set, (b) selecting a sample and the variables to study, and (c) assessing data reliability along with stating the specific questions to be answered. Once these steps were completed, the data analysis followed the same guidelines as a primary analysis.
The data were selected from the open-access LatamChequea Coronavirus network database. This is a database nurtured by members of the fact-checking organization network. Once a claim was identified by an organization, it was registered on the collaborative database and analyzed according to the set of variables. The database contained five types of verified content: viral disinformation, debunking public figures, debunking media, explanatory pieces, and other content (everything not included in the previous categories, such as content published on an official government or health organization website). In this research, we analyzed only the COVID-19 viral disinformation (n = 734) fact-checked from January 18, 2020 to April 24, 2020.
The assessment of reliability was carried out following Riedel’s (2014) method. The database was analyzed according to the LatamChequea Coronavirus codebook.
Quantitative Content Analysis
A total of 734 viral disinformation pieces in Spanish and Portuguese performed by 25 Ibero-American organizations were analyzed (Table 1).
Volume of Disinformation Collected in the LatamChequea Coronavirus Database.
The variables studied from the LatamChequea Coronavirus database are: name and country of the organization that did the fact-checking, title (a description to identify the disinformation), final rating (assessment of the content), date of the check, information type/topic (i.e., the situation of the country, government measures, prevention, origin of the virus, symptoms, characteristics of the disease, cures, contagion, and others), format of the content (i.e., audio, still image, text, voice, and video; memes were not fact-checked), origin of the content (i.e., Facebook, WhatsApp, Twitter, Instagram, YouTube, social networks, online and traditional media, SMS), and whether the disinformation circulated in other countries. Before verifying a claim, fact-checkers confirmed whether this disinformation was in the database. If it had already been debunked by another organization, fact-checkers recorded that it had also circulated in their country.
The variable “final rating” is how fact-checkers labeled the assessment of the content. Each organization has its typology, but some labels are common. In total, we identified 22 labels and regrouped them into clusters with similar definitions, according to each organization’s methodology published on its website. For example, the labels “true” and “correct” were subsumed into the label “true,” and the labels “accurate but” and “true but” into the label “true but.”
The set of evaluations was reorganized into 11 labels: false, misleading (questionable or doubtful), out of context, manipulated, unsustainable (any content where the source cannot be identified), no certainty, partially false, true but (reliable information rendered invalid by the omission of data relevant to its proper understanding), multiple checks, hasty (the situation is ongoing, so that, any conclusions are therefore preliminary), and satire. The quantitative data analysis was done with Python.
Second Phase: Primary Analysis, In-Depth Interviews
In the qualitative part of the study, we performed in-depth interviews (n = 12) with fact-checkers from the LatamChequea Coronavirus network and the deputy director of the IFCN—the organization that supports LatamChequea—who provided an international perspective about the collaborative work in the region.
We conducted online interviews that lasted an average of 1 hour between April 10 and May 4, 2020. They were recorded on video, transcribed, and translated into English for analysis. The transcripts were examined by assembling thematic blocks and identifying emerging categories.
After requesting the participation of the network members, fact-checkers from nine different countries and 11 organizations (Table 2) agreed to respond to our study. The questionnaire (see Appendix 1) was intended to examine the experiences of fact-checkers on (a) assessment of the COVID-19 disinformation coverage, (b) challenges brought by the coronavirus pandemic, (c) solutions to fight coronavirus, and (d) analysis of transnational alliances.
Interviews With the Ibero-American Fact-Checking Organizations.
Note. IFCN = International Fact-Checking Network.
Results
Results of the Secondary Content Analysis
Most of the disinformation registered in the LatamChequea Coronavirus database was evaluated as false (92%), for example, the claim that coffee cured COVID-19. That hoax circulated in Colombia (debunked by Colombia Check on April 10, 2020), Mexico (debunked by El Sabueso on April 2), Costa Rica (debunked by #Nocomacuento on April 3), Venezuela (debunked by Efecto Cocuyo on March 30), and Spain (debunked by Maldita.es on March 13 and again on April 2).
To a lesser extent, the fact-checked content was classified as misleading (3%) or as information taken out of context (1%). A tiny proportion was labeled as manipulated information (0.7%), such as a video of Vladimir Putin, the Russian president, in which he allegedly claimed that world leaders wanted to reduce the population through COVID-19. El Sabueso (Mexico) debunked this video on March 19, identifying it as a speech given by Putin in 2016 whose subtitles had been manipulated. Residually, the fact-checked content was classified as unsustainable (0.4%), no certainty (0.3%), partially false (0.3%), true but (0.3%), multiple checks (0.3%), hasty (0.1%), and satire (0.1%).
Regarding the topics of COVID-19 disinformation, the majority were registered under the category “other” (25%) which contained all the topics that were not included in other categories but were related to the health crisis. For instance, a hoax claimed that Netflix would give free access to its catalog during the lockdown. That claim circulated in Spain (debunked by Newtral on March 23), Brazil (debunked by Lupa on March 19 and Estadão Verifica on March 20), Bolivia (debunked by Bolivia Verifica on March 20), Paraguay (debunked by El Surtidor on March 24), and Mexico (debunked by El Sabueso on April 2). This example shows the capacity of hoax virality, in which the same disinformation spread in less than 2 weeks in five different countries.
Apart from the container category “other,” the second most prevalent topic was the measures taken by governments (18%), followed by the country situation (15%). The content about false cures of COVID-19 (11%) is in fourth place, followed by means of prevention (8%), the origin of the virus (6%), and contagion (5.4%).
The disinformation evaluated by the LatamChequea Coronavirus network circulated mainly on social media, such as WhatsApp (28%), Facebook (20%), generic social networks (16%), Twitter (9%), Facebook, and WhatsApp together (7.6%), and simultaneously through two (1.6%) and three (0.1%) social networks.
The format in which most disinformation was shared during the first three months of the pandemic was text (46%), followed by video (16%), still image and video (15%), only still image (10%), and audio (6%). The rest of the fact-checked content involved more than two formats (2.1%). For instance, in Spain, Maldita.es (February 13th) debunked a photograph of four people cutting open the corpse of a baby with the caption: “Now they also eat babies in Asia.” The image, which circulated on Instagram, corresponded to an autopsy on a baby in a forensic facility in Bangkok.
Fact-Checked Content Flows and Lifespan in Ibero-America
To know the lifespan of COVID-19 fact-checked content, we performed a multivariable analysis between the “information type/topic” and the “date of the check” to observe both behaviors across time.
The highest quantity of COVID-19-related disinformation was fact-checked between March 16 and 29 (26%), when half of the Ibero-American countries represented on the LatamChequea Coronavirus database went into lockdown (Spain, Ecuador, Brazil, Costa Rica, Venezuela, Chile, Peru, and Uruguay). This happened 2 weeks after the WHO announcement classifying COVID-19 as a pandemic. The other half entered lockdown between March 20 and 31 (Colombia, Mexico, Argentina, Bolivia, Paraguay, Cuba, Guatemala, and the Dominican Republic), except for Nicaragua, which did not apply lockdowns during the months analyzed.
The peak of fact-checked content coincided with the most critical phases of the evolution of the first wave in Spain, even though the rest of the countries were in an escalation phase.
When disaggregating the data per fortnight, as shown in the histogram (Figure 1), the topics “cures” and “prevention” maintain approximately the same number of occurrences from the second week of March to almost the first 2 weeks of April, when the number of claims began to drop sharply.
Histogram of Fact-Checked Topics per Fortnight.
This decrease in viral disinformation could be explained by the measures implemented by WhatsApp to reduce the spread of hoaxes. In 2018, WhatsApp limited the forwarding of a message to five chats, which caused a 25% decrease in message forwarding worldwide. In 2019, the platform identified with a double arrow the messages that had already been forwarded. During the pandemic, it restricted the forwarding of “highly forwarded” messages, thus reducing sharing by 70% (Singh, 2020). While the drastic reduction in disinformation cannot be solely attributed to the latest WhatsApp measures, these undoubtedly hindered the dissemination of hoaxes.
The topic “government measures” had more occurrences during the last 2 weeks of March, corresponding to the beginning of the lockdown in half of the countries in the sample.
The bivariate analysis of the “country of the organization that did the fact-checking” and “the disinformation circulated in other countries” allowed us to track disinformation flows among countries from a contingency table, shown in the following heatmap (Figure 2).
Heatmap of Disinformation Flows Among Countries.
The heatmap illustrates that the highest flow of disinformation registered by the fact-checkers occurred between Colombia and Spain. Significant flows are observed among Spain, Mexico, Argentina, and Colombia. The same disinformation circulated from one country to another at an accelerated rate. Occasionally, the same disinformation spread in several Ibero-American countries in less than a week.
These results highlight a political crisis frame due to the prevalence of disinformation related to governance/state categories (measures, country situation, etc.) rather than just treatments or prevention.
Results of the In-Depth Interviews
Assessment of the Coverage: The Political and Panic-Provoking Intentions
Confirming the content analysis results, the fact-checkers identified diverse disinformation topics, such as government measures, cures, disease characteristics, forms of contagion, vaccines, treatments, and so on. According to the interviewees, in the very first phase of the outbreak, the most widely disseminated COVID-19 disinformation referred to home cures and the advice of so-called experts. The fact-checkers considered these types of disinformation dangerous, since they could produce a false sense of security in the population, prompting people to reduce or dispense with personal protection measures: There is a [fake cure disinformation] that poisoned a lot of people in Brazil. It was putting a shot of bleach in a glass of water. [It said] If you drank it, you will be cured of the coronavirus or avoid being infected! (Director 3, Agencia Lupa-Brazil)
Moreover, participants revealed other critical disinformation topics focused on destabilizing governments, the press, or public entities. These included, for example, disinformation related to the measures promoted by the WHO, scientific studies on the virus, expert advice, the supremacy of some race and/or religion, and conspiracy theories about the origin and causes, among others. Occasionally, these conspiracy theories and disinformation were spread by public figures and politicians.
As reported by fact-checkers, the information disorder in Ibero-America—principally disinformation—had political and panic-provoking intentions. Concerning the disinformation they received, Journalist 4 from Gato Encerrado-Despacho 505-Nicaragua put it like this: “In Nicaragua, . . . it is rather a health crisis with political overtones.” Similarly, Director 6 from Agencia Ocote-Guatemala indicated that there is “. . . a lot of panic generation, but it is also largely a product of the opacity and lack of transparency of the [Guatemalan] government.”
Director 2 from Bolivia Verifica stated that disinformation was used as a purely political strategy. He explained,
The particular problem in Bolivia is that the pandemic crisis has been mixed with the political crisis that we already have, and political actors are using the pandemic to create destabilization. We have a large amount of disinformation generated by the political crisis and the pandemic . . . and citizens are receiving ambiguous messages from all sides.
Director 3 from Agencia Lupa-Brazil pointed out the clear use of disinformation by Brazilian politicians to undermine informative media: It is a difficult time for the press in Brazil because we have a government that works with disinformation as a political strategy, so one of the disinformation strategies is to discredit the press. It [the government] says that the press is the enemy and journalists are trying to affect the government. It is a way to minimize the work of the press and attack the press as an institution.
On the other hand, Verificado and El Sabueso in Mexico perceived that one of the aims of disinformation in that country was the destabilization of the state, especially concerning the measures adopted by the government to counter the pandemic. The spread of disinformation is used “as a political argument for discrediting the actions of governments both in Mexico and in other countries” (Journalist 2, El Sabueso-Animal político-Mexico): The disinformation in Mexico is related to the lack of understanding of the official information. It is related to the premise that the government is not doing anything. It has a very important political component where powerful groups are constantly attacking the decisions taken by the federal government or each federative entity. (Director 7, Verificado-Mexico)
The qualitative results reveal the politicization of the pandemic and how COVID-19-related disinformation represents a political crisis in some Ibero-American countries according to the interviewees’ perceptions. The fact-checkers claimed they did not wish to be regarded as replicators of government information. Consequently, they highlighted the crucial role of journalists as explainers in contextualizing and contrasting information provided by official sources and making it useful for the public. Respondents see contrasting and explaining journalism as a way to stop the waves of panic-causing disinformation, regardless of whether they are generated by the state or private interests.
Perceptions of the Sources and Formats of COVID-19 Disinformation
The participants stated that the dominant sources of COVID-19 viral disinformation in Latin America are social networks: Facebook, Instagram, Twitter, YouTube, WhatsApp, and Messenger. According to the fact-checker’s perceptions, Facebook and WhatsApp lead this category, tied to the quantitative results obtained.
As Director 8 from the IFCN affirmed, “Latin America is very strong in WhatsApp, unlike the United States or Asia.” This differentiation is determined by the social context of each country along with the internet consumption habits of its citizens. According to the perception of Director 4 from Colombia Check, Facebook and WhatsApp are the niches for spreading hoaxes in Colombia: Most people who have access to the Internet do so through their smartphones. In most mobile phone plans, they have unlimited use of Facebook or WhatsApp but cannot access Google or [sites like] Colombia Check. . . . This is one of the reasons why both information and disinformation move around there [social media]. (Director 4, Colombia Check)
The perception of Director 2 from Bolivia Verifica is that “COVID-19 disinformation in Bolivia circulates mainly through Facebook and Whatsapp,” the most used social networks in that country. Likewise, Director 5 from Ecuador Chequea affirms that in Ecuador, “WhatsApp is the main dissemination platform of disinformation, followed by Facebook.”
One of the recurrent obstacles for fact-checkers is the source identification of the hoaxes (creators) due to the privacy policies in force on social networking platforms. Journalist 1 from El Detector de mentiras-La Silla Vacía-Colombia commented that despite it being arduous to find out who designed and shared the content, in some cases, they were able to . . . “identify as a source of disinformation influencers on Twitter who are part of polarized political communities, in addition to specific web pages created with the purpose of disinforming”: There is disinformation of many kinds. However, there is a constant type that has grown exponentially. It is the chain hoax messages on WhatsApp. They are the most difficult to trace, which are the worst. (Director 7, Verificado-Mexico)
Some network members explained that not only the virus mutates, but also disinformation strategies and formats. For example, in Mexico, a hoax mutated from a fact-check: A media that spreads false news took a debunking we did [El Sabueso] and changed the “no” to “yes” so our verification work was used to sustain a piece of false information. (Journalist 3, El Sabueso-Animal político-Mexico)
Traditional media is also fact-checked. The interviewees commented that most of the disinformation found in media came in the form of research papers that were in a testing phase or had not yet undergone peer review. In those cases, the information piece was easily rectified. “If we [Agencia Lupa] see something wrong in traditional media, we call them, point out what is not true, and suggest it is better to change it on their website” (Director 3, Agencia Lupa-Brasil).
A finding from the interviews is that, although these fact-checking organizations are not dependent on traditional media, the fact-checkers’ intention is to maintain open channels with other journalists. “This moment [the pandemic] invites solidarity among the media. The [journalistic] teams on the street covering the information are teams that are helping each other” (Director 2, Bolivia Verifica). The interest in this mutual collaboration and solidarity in a united front came at a time when “the speeches of authoritarian governments [were] discrediting the work of journalism” (Director 5, Ecuador Chequea).
Challenges in the Verification Process During the Outbreak
Although the structural protocols and methodology of fact-checking have not changed with the coronavirus outbreak, diverse factors have affected the verification process. One of them is the fact-checkers’ skepticism about the information provided by governments due to the uncertain reliability of official sources: . . . Let’s say we have to do a double, triple fact-check on what someone from the Ministry of Health officially tells us because we are not sure that the information the ministry spokesperson is giving us is real. (Director 6, Agencia Ocote-Guatemala)
In the case of Nicaragua, the press was not able to provide the public with helpful information due to the secrecy of government sources that minimized the health crisis. This also led to organizations needing more time to carry out their fact-checks. In the words of Journalist 4: The Health Minister offers press conferences but only the public media have access. There is no chance for questions, much needed at a time like this one. The First Lady, who speaks every day in Nicaragua, is not clear with the information she provides. Verification of what they say is necessary and takes. time (Gato Encerrado-Despacho 505-Nicaragua)
The situation in Bolivia was very similar to what was happening in Guatemala and Nicaragua, as stated by Director 2: There is little trust in state institutions because they have not earned it. People want to know what will happen, how they will be cured, and who will treat them if they get sick. Unfortunately, the government does not know how to give this information. There are permanent contradictions and a lack of coordination in their different entities. (Bolivia Verifica)
In this way, distrust in institutional sources constitutes a challenge that fact-checkers must face. Challenges are determined by the political environment where each organization is located (Amazeen, 2019).
Other factors that have influenced the verification protocols were the specialized nature of the information (scientific and medical) and its timescale, as it is updated day by day: Before, the verification process was simpler. Now, as it is more scientific, verdicts cannot be given so quickly as many specialized sources have to be consulted. (Journalist 1, Detector de mentiras-La Silla Vacía-Colombia)
It has been a challenge for the fact-checkers interviewed to undertake the verification of health issues as well as find reliable scientific sources: It is the first time since I began doing journalism, ten years ago, that science has such a determining role. And I think that we journalists have had very little involvement with science, which has prevented us from making sufficiently precise interpretations [on health information]. (Director 5, Ecuador Chequea)
Finally, over-information has been another decisive factor in the verification process. As stated by the interviewees, information overload on the pandemic has displaced other issues on the public agenda that become no longer relevant. The coronavirus slowed the spread of other types of circulating false information.
Mapping the Solutions to Fight the Infodemic
During the health crisis, the Ibero-American fact-checking organizations implemented various strategies to combat disinformation and reach a larger audience. Fact-checkers developed different approaches to transmit accurate information by adapting to their context.
Agencia Lupa-Brazil created a database of COVID-19 scientific studies translated into simple language, a podcast in which an infectious disease specialist answers listeners’ questions, and an artificial intelligence software that connects the population with reliable data and sources. Moreover, the agency in alliance with the Redes Cordiais initiative used influencers to disseminate verified information on the virus. These influencers created awareness of disinformation risks through live broadcasts.
La Silla Vacía-Colombia focused on improving the interaction with the audience. Its site has an interactive section called “Sea usted el detector” (You be the detector) in which readers participate in a quiz to discover if they guessed the correct rating given to fact-checked content. Therefore, users could contrast the information they received. Colombia Check emphasized producing explanation pieces in plain language, in text, and in audio. They also created a special coverage section on the virus on the website.
Verificado-Mexico provided explanatory content through podcasts and texts, along with training activities for journalists. Animal Político-Mexico opened on its site a special section for news, fact-checks, graphics with statistical data, interviews with experts, and explainers in video and text.
Ecuador Chequea, which works in collaboration with academia, journalists, and the public, produced explanation pieces in Spanish and indigenous languages in text and video format to reach minority communities.
Agencia Ocote-Guatemala carried out campaigns to encourage the audience to send information to be fact-checked. It also offered podcasts tied to journalistic projects and special coronavirus coverage that combined its own and shared content from other organizations in the network.
Bolivia Verifica, besides publishing fact-checking pieces, produced informational videos, podcasts, newsletters, and digital maps with local, national, and regional data.
Transnational Alliances: The LatamChequea Coronavirus Network
Regarding the alliance, all the journalists interviewed highly valued the collaborative work and communication among the fact-checkers of the region. They believed that belonging to the network and being endorsed by the IFCN and the Poynter Institute enhanced their credibility in the eyes of their audiences.
Although the network was created long before the outbreak, one of the strongest and most necessary collaborations has emerged from the creation of the shared regional database in Spanish and Portuguese on coronavirus disinformation. “LatamChequea does not compete with but complements the work of the IFCN, whose primary language is English. . . . It is an example of how regional collaborations can also be very powerful” (Director 1, Chequeado-Argentina).
Undoubtedly, for the interviewees, what stands out the most is the trust in the rigor of the work done by the other fact-checkers. Also, they appreciate the ability to access and share verified information with their audiences, not only in Ibero-America but throughout the world, thanks to the Coronavirus Facts-Alliance: Each fact-checker does a check on their country and contributes with that information to a central database [the Coronavirus Facts-Alliance database] so that others who have an interest can reproduce it. 88 organizations in 74 countries, with content in 43 languages. We are a very small army to combat all the disinformation in the world. (Director 8, IFCN-United States)
During the pandemic, these alliances have provided fact-checkers with valuable clues about the impact of their work in Ibero-America. There are no officially sanctioned indicators measuring the impact of their fact-checking, debunking, and explanatory content. However, the organizations in the sample have seen exponential growth in the number of website visits, the size of their social network audience, and engagement—in the form of fact-check retweets and shares. Furthermore, health institutions have disseminated information published by fact-checkers: What I see is that the audience is growing and our content is being spread on social media. People are answering messages on social networks with the fact-check made by Lupa, using Lupa as a source. (Director 3, Agencia Lupa-Brazil)
While it is possible to have an idea of the extent to which fact-checks are shared on social media, there are still limitations. For instance, this type of information is not available on WhatsApp. Nevertheless, the increase in the audience has shown that during the first stages of the outbreak, people started to be more aware of the information they received and questioned its veracity. Regardless of any impact that fact-checking organizations might be able to measure, they agree that their first concern is to provide accurate information to the public, so that, they can make conscious informed decisions.
Discussion
This exploratory study examined how fact-checkers understand the information disorder in Ibero-America, particularly COVID-19 disinformation, and what strategies they employed to address it in crisis contexts.
To understand this information disorder, we identified the prevalent interpretations (frames) and characteristics of the COVID-19-related disinformation. The content analysis results showed that most fact-checked content was incidental issues with a minimum connection to the virus, generating noise and over-information during the first 3 months of the pandemic.
The measures taken by governments were a prevalent topic of COVID-19 disinformation. This added to the constant erosion of trust in institutions, pointing to the observations of Papapicco (2020) and Cinelli et al. (2020), who warned that disinformation undermines the effectiveness of public health campaigns. Thus, this verifies that the COVID-19 disinformation registered in the LatamChequea Coronavirus database had a relevant political intention, as shown by the results of López-Pujalte and Nuño-Moral (2020), Salaverría et al. (2020), Sánchez-Duarte and Magallón Rosa (2020), and Gehrke and Benetti (2021).
The qualitative interviews framed by the quantitative content analysis revealed the politicization of the COVID-19 disinformation. The results from both techniques showed that this disinformation should be seen through the lens of a political crisis due to the prevalence of disinformation related to governance, as well as the fact that fact-checkers believe that COVID-19 disinformation must be understood in the context of political discourse and strategy.
Our findings expose, according to the fact-checkers’ perceptions, that the health crisis in Ibero-America turned into a political crisis due to factors, such as the lack of transparency of governments as information sources, politicians, and public figures spreading information with no scientific basis, disinformation generated to discredit the leadership of governments and institutions, and finally, politicians discrediting the press.
The COVID-19 disinformation registered by the fact-checkers had a two-way flow among Spain, Mexico, Colombia, and Argentina. The highest concentration of flows was between Spain and Colombia. This phenomenon might be explained by the presence of a large Colombian diaspora in Spain, a potential subject for further research.
The findings of this research reinforce findings from Cinelli et al. (2020), Li et al. (2020), Papapicco (2020), and Rubin (2019) that the main channel of distribution of disinformation is social media. Most of the COVID-19 disinformation found by fact-checkers went viral in Ibero-America on Facebook and WhatsApp. Twitter carried the least. Some telecommunication companies in Latin America include Facebook and WhatsApp in the basic rate plans, which explains their widespread use. While hoaxes are disseminated faster on social media, the verification process takes time. As the content analysis showed, a hoax circulated in seven different countries in a single week. In response, fact-checkers call for greater control by social media platforms to tackle the infodemic.
Before the coronavirus outbreak, some associations only had a political emphasis as they only fact-checked public figures’ and politicians’ claims. However, we found that the Ibero-American fact-checking organizations in the sample also fact-checked viral disinformation and political discourses, trained journalists and the public in media literacy, and mainly in this context of crisis, produced explanatory health, political, and economic pieces.
The results of the present study allow us to advance knowledge in fact-checking practices, as recommended (Dan, 2021). In the Ibero-American context, fact-checkers have shown how the use of multiple strategies to communicate messages (checks and explainers) could be effective in reaching their audiences. These include, for instance, the use of autochthonous languages for indigenous communities; formats like illustrations, videos, or podcasts for the population unable to read or write; influencers to interest younger generations; and interactive content to engage wider audiences.
Other lessons learned from the Ibero-American case regarding fact-checking practices include the double contrast of information from official sources—characterized by their opacity, the need for constant training of fact-checkers on all kinds of issues, and above all, the transnational fact-checking alliances as a badge of trust in the rigor of other fact-checkers and the possibility to share information regionally.
When we consider that many of the organizations analyzed are digital media without national boundaries (Higgins Joyce & Harlow, 2020; Vu et al., 2023), it becomes clear that the Ibero-American fact-checking associations are an example of how transnational online journalism contributes to strengthening democracy.
Although they are challenged with limited financial resources, the participants expressed that their organizations have found in the collaborative work (Cueva Chacón & Saldaña, 2020; Hellmueller et al., 2019; Stonbely, 2017) of the LatamChequea network an effective means of reinforcing citizen participation in public debates. Fact-checking alliances tend to rise in democracies under risk (Amazeen, 2019). That is why the LatamChequea Coronavirus network represents a milestone in collaborative journalism throughout the region.
Limitations
There are some limitations to this study. The first limitation is the fact that the analysis of disinformation is based only on the fact-checked content registered in the LatamChequea Coronavirus database. Despite fact-checkers’ efforts in verifying claims, they chose what sources and content to fact-check according to their work overload and the personnel in each organization. Those aspects could affect the number of fact-checks the associations provided to the database. Therefore, the total of checks collected is not representative. Also, the results we provide should be understood considering the context and the method used in this research. Although the database included fact-checks in Spanish and Portuguese, there were language barriers when contrasting and sharing the checks as not all fact-checkers could understand both languages. Finally, the period of the data set is sufficient to answer our research questions, but does not allow a broader observation of the fact-checked content behavior.
Footnotes
Appendix
Acknowledgements
The researchers would like to thank the IFCN for granting full access to the CoronaVirusFacts Alliance database. They also thank the journalists’ members of the fact-checking organizations that make up the LatamChequea Coronavirus network for collaborating with this research.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
