Fake news and Disinformation in Roma Communities – research summary

Background of the research

The main objective of the RADIRIGHTS project is to identify and combat disinformation affecting Roma communities. In order to understand the mechanisms of disinformation, we needed to identify and learn about the disinformation circulating in the Roma communities in the partner countries (Bulgaria, Czech Republic, Spain and Hungary)

To this end, between November 2023 and February 2024, partner organisations interviewed community members in all project sites, in a total of 12 locations, to find out what news or social media stories they had encountered before or during this period that influenced them. These cases of disinformation were collected and analysed using a set of eight criteria. The collected cases will be analysed in detail during the summer workshop 2024.

The work was complemented by desk research in Germany by a German native speaker who was an intern of the Autonómia Foundation, who was asked not only to collect disinformation stories in Germany, but also to analyse disinformation cases in German public life and to collect the countermeasures taken by the German federal government.

The data collection was further complemented by extensive information and data collection on social media platforms in Hungary, including a review of government channels and Facebook groups established to influence public opinion as well as a search for disinformation incidents. We also held workshops with secondary school students in four locations across the country addressing this issue. Although this age group is not a target group for the project, we found it very useful to involve them in the work, as their activity on social media is different and more active than that of older age groups. We also felt it was important to target this group because it is easier to prepare them to recognise and act against disinformation content.

As part of the project, the Eötvös Lóránd University of Applied Sciences organised a research seminar with invited Roma participants, followed by a university lecture on the seminar’s experiences.

Using the experiences from the university and fieldwork, a one-day workshop will be held at one of the most popular festivals in the country (www.kolorado.hu), involving Roma and young people attending the concert, to identify and playfully practice countering disinformation.

Although these activities are not included as deliverables in the project description, they are integral to the project content and indicate that the results of the project will be put into practice during the project period.

Disinformation cases

The collected cases of disinformation were grouped into different categories. The first grouping criterion is whether the disinformation case is harmless or harmful to the Roma community. This was decided subjectively by the respondents on the basis of whether the incident itself had an impact on the local community and, if so, what impact.

The next, related categorisation criterion was whether the harm caused by the spread of the disinformation was intentional or without a manipulative background, i.e., whether it tended to fall more into the so-called misinformation category. If intentionality or manipulative intent can be clearly identified, then we must identify who the actors are that play a key role in the creation or spread of such disinformation.

We have also attempted to identify the real intentions behind the spread of disinformation. Is the news based on targeting those of a different socio-cultural background, or is there a broader, political background to the spread of disinformation?

Perhaps the most important aspect of this analysis, however, was whether the disinformation case identified was Roma-specific or whether it was specific to society as a whole. Is it being exclusively or just partly spread within the Roma community, is its origin linked to this community, or does it have no ethnic background whatsoever? Such disinformation may be specific to one country or may appear in several countries in a similar way.

When analysing cases of disinformation, we also considered whether the case could be used to reinforce stereotypes about Roma, especially those related to their undereducation and manipulability.

Another analytical aspect was the way in which the disinformation content is spread, i.e. whether it is linked to the consumer’s educational or cognitive level or – as we have seen mainly in the case of certain conspiracy theories – whether it is also at least as present among citizens with higher education.

One example of such a conspiracy theory is the chem-trail theory, for example, or theories of the world that attribute political influence to various secret groups (a good example of the latter is the idea, promoted by the Hungarian government, that the figure of George Soros lies behind practically every negative social phenomenon in Hungary).

A separate category of disinformation cases is that of content related to social situations, which is given currency because of the social vulnerability of those purportedly involved. Specifically, this is misinformation related to the state or municipal benefit system, which often increases dissatisfaction with the benefit system and, although it is not usually a deliberate manipulative tendency, activates the audience and thus further undermines faith and trust in social institutions. More generally, this has to do with the relationship between the majority society and Roma communities.

The cases 

Of the cases collected, it is difficult to detect manipulative intent in most of them. Although we avoided collecting disinformation cases with direct political content in Hungary, as this would have made the implementation of the project (or more precisely the functioning of the project host organisation) even more impossible, it can be said that the Roma community is at least as targeted with political disinformation as is the majority society as a whole. Moreover, in the case of Roma communities, their access to different sources of information is even more limited for geographical and social reasons, which makes the manipulation much more effective in their case.

Political disinformation was also reported by our Spanish and Czech partners, but was more likely to be local government-related and on a larger scale than the general disinformation that might influence society as a whole.

The data collection itself took place a year and a half after the last COVID closure, but we still found anti-vaccine content and theories outlining conspiracy theories as to what was behind the spread of COVID to be present online. On the whole, however, it can be said that these communities have in fact moved on from COVID as a subject, and at the time of data collection, the issue was no longer a major theme.

Anti-vaccine views, which were directly addressed in the call for proposals (CERV), therefore were not included in the focus of the project. Rather, the focus shifted to the social issues that also characterised the lives of these communities prior to the COVID pandemic.

Half of the Hungarian cases are linked to government propaganda. Disinformation content related to refugee or LGBTQ communities has regularly appeared in Hungarian public media and government-supported social media since 2015. In two of our locations we encountered mentions of this content, or more specifically, the views that have emerged as a result of this exclusionary content, present in the life of the local community.

In one of the sites, in a very disadvantaged small village (with a homogeneous Roma population), Roma women reported that they were afraid of migrants who were violent towards women and associated them with other crimes. 

Most of these women said they had never personally encountered migrants in their lives. Those who had met migrants had not been subjected to violent crime by them. 

Migrants are generally perceived by members of this group as dangerous people who intrude into the life of a society. This generalised, exclusively negative image is also strange because the Roma say their own children and close family members in Western European countries are in a similar economic migration situation. Yet they uniformly endow other migrants with characteristics conveyed by the media that are not borne out by their experiences in Hungary, and they get their information exclusively from government-controlled media.

The Roma community is clearly marginalised in Hungary, and people belonging to this group also become marginalised if their opinions are shaped by one-sided information alone, with no experience to either refute or support it.

A similar pattern was observed in another municipality, where young adults reported that the portrayal of gay or LGBTQ communities in the media posed the risk that viewers of such content could change their sexual orientation. In addition, members of this group view non-heterosexual people in a judgmental, aggressive manner, completely rejecting equal rights for them. Their information comes mainly from the media, they have no personal experience with this group.

No political motivation can be identified behind the content of several other kinds of disinformation. This is not the case, for example, with cases related to consumption patterns specific to extreme poverty. Expensive smartphones are highly prestigious items, especially among young people in urban areas. A common misconception, reinforced by commercial advertising and in many cases by the manipulative communications of telecoms companies, is that it is not necessary to pay the full purchase price for such a device, and in fact the full price is not even clearly disclosed. The loan products offered for procuring this technology show just the cost of the first down payment, with the total purchase price barely visible in the advertisement. Thus, buyers are led to believe that a smartphone costing several hundred thousand forints can be obtained for a few tens of thousands of forints. Their inability to pay the full price leads to their subsequent indebtedness and loss of future creditworthiness. 

A further typical area of disinformation is in the range of health and medical issues, which is driven by a lack of knowledge and the dysfunctionality of the Hungarian healthcare system. In the case of an emerging illness, the Roma respondents do not expect to access specialist care, but start self-medicating on the basis of misconceptions spread by social media. The social damage of this cannot be underestimated, because by the time such persons do access specialist care, in many cases it is much more difficult or impossible to cure their illness.

An ideal case of this type of disinformation is the fake news claim that sodium bicarbonate can cure cancer. In our previous research on the relationship between Roma and the health care system, we found several cases of the non-indicative use of sodium bicarbonate for digestive problems.

In Spain, a number of disinformation reports about the time of the COVID outbreak were identified which were based on half-truths or incorrect information and for which it was difficult or impossible to identify the source of the information.

During the COVID outbreak, local authorities were unable or inadequately equipped to communicate the rules to be followed during the outbreak, leading to misunderstandings. These cases were about the disproportionate, unlawful use of official measures against Roma communities.

These cases lead to Roma communities losing trust in local authorities and not wanting to cooperate with them further.

In two cases, a very similar pattern was observed: A rumour spread that the municipal social welfare system was taking children away from low-status poor families, removing them from their families. In another case, colleagues in Roma communities encountered a rumour that Moroccan immigrants receive much higher social benefits than Roma families with similar social status. This, according to local colleagues, has no basis in reality, but it also has the potential (as we saw in the Hungarian case) to pit different minorities against each other and ultimately to reinforce the identity of a community, influenced by disinformation, at the expense of another excluded group.

In two other cases, the Spanish colleagues found examples of stereotypical images of Roma. One is a distorted, untrue representation of the employment rates of young Roma, while in another case, an unscientific research report tries to justify the predominance of Roma men as a result of masculine culture, violent behaviour and a desire for domination. In both cases, the news was published through a high-circulation press product. 

The following case described by the Spanish colleagues is also present in all partner countries: The claim that NGOs working for Roma inclusion use the available public funding resources for their own purposes, i.e., that they steal money that could be going to aid the Roma. There is usually no data-based evidence behind such accusations and these rumours are a way of discrediting not just Roma, but also pro-Roma or Roma-run NGOs, thereby further destroying the already not very good relations between Roma and non-Roma society.

The Czech cases have focused on anti-Ukrainian sentiment increasing among Roma as a result of media coverage of altercations between members of these communities and the spreading of rumors on social media. Many Ukrainians fled their country as a consequence of the Russian-Ukrainian war and have become a prevalent minority in recent years.

In the Czech Republic, tensions between ethnic Ukrainians and Roma in the Czech Republic escalated in response to an altercation in June 2023 during which a Roma man was killed by a Ukrainian man (who was subsequently acquitted in March 2024 for having acted in self-defense).

This incident became the basis for some Roma to generalize about the motivation of this and other violence in order to implicate the entire Ukrainian minority as co-responsible for attacks on Roma. Behind this there are general anti-Ukrainian and presumably also Russian propaganda efforts.

The conflict between the Roma and Ukrainian minorities has given rise to regular outbreaks of interpersonal violence. In addition to actual incidents of violence against Roma which have triggered commentary and panic on social media, some Roma social media users have also deliberately published misleading claims about non-existent incidents allegedly committed by Ukrainians. 

However, unlike the cases from other countries, in the Czech Republic police are investigating these claims and will prosecute social media users for spreading false news if that can be proven.

The practice of the Czech media, police and politicians is not uniform when it comes to disclosing the citizenship, ethnicity or nationality of persons allegedly involved in criminal activities. In the deadly incident in June 2023, the police officially identified the perpetrator only as a “foreign nationals”, leading many Roma people who had heard the perpetrator was Ukrainian to suspect that Ukrainians were being given special protection by the authorities. Even in cases where the police do not disclose such information, local politicians may refer to it in their comments to the press. In sloppily-compiled media reports, no further information about the suspects is provided (such as how long they have been living in the Czech Republic), which could hamper further speculation. This reinforces the assumption that all the Ukrainians (or “foreign nationals”) involved in such incidents are coming directly from the war and committing violence in the Czech Republic, when the reality is more complex (there has been a large Ukrainian diaspora in the Czech Republic for decades). The information vacuum is filled by rumours and speculation. 

The pattern that appeared in the Spanish example is also present in the Czech one: A meme based on a photograph from a Roma demonstration was used to allege that Roma are “welfare cheats” and indirectly to deteriorate the relationship between Roma and non-Roma communities (the image shows a Roma man holding a beer can, taken from a video in which he is criticising the quality of social benefits).

While in Spain, “news” of official restrictions on moving during COVID were actually a hoax, in Bulgaria, access to and from Roma neighbourhoods was in fact restricted in some cities where infection levels were reported to be too high. The media attributed these decisions to Romani people’s alleged failure to comply with the measures in place. In fact, during the curfew period, rumours appeared in mainstream media outlets with large audiences that the infection was mainly being spread by members of Roma communities. Reports began to circulate in the media that Roma were failing to comply with disease control measures and were leaving their homes when they were not supposed to, which is thought to have greatly accelerated the spread of the virus. 

Another fake news story in Spain claimed that children taken from Roma families were being offered for adoption to gay families in Norway. This hoax has two damaging effects: It undermines confidence in the social care system and it is also anti-gay.

Spreading through Facebook, a typical conspiracy theory was that the main reason for the COVID epidemic was that “Bill Gates is spying on us!” The conspiracy theory was that the virus had artificially been created by Bill Gates, whose main goal was to then develop a vaccine against it. This vaccine would be used, according to the fake news, to implant tracking devices in the form of microchips under people’s skin and to collect information on the behaviour of these unsuspecting, vaccinated citizens. This conspiracy theory is not specific to Bulgaria, but is particularly popular among Bulgarian Facebook users. 

Summary

A clear anti-Roma motive can be found in only some of these cases, but the secondary responsibility, the inaction of those in political power, must also be mentioned. Exept the Czech Republic, no state action against disinformation was found in the countries studied. In Hungary and the Czech Republic there are initiatives in the independent media (Lakmusz, Romea.cz). 

In Hungary, paradoxically, trust in the public media is behind the effect of disinformation or at least of one-sided information. Instead of playing a major role in the fight against disinformation manipulation, the public broadcast media (which are indistinguishable from the pro-government satellite media in terms of content) tend to amplify, or at least not to refute, news that is typically intended to strengthen the position of the Fidesz government, whose manipulative intentions are difficult to dispute.

The RADIRIGHTS project does not aim to collect and analyse stereotypical content about Roma in the media, but other research has already done a lot of work on this issue. What is new in this project, in the collection of disinformation content, is documentation of the negative attitude towards Roma and other minorities embodied by this content, the rejection of minorities presented in a similarly stereotypical way.

Being a member of the Roma minority does not reinforce a sense of solidarity with other minorities. On the contrary, it is possible for members of the Roma minority to reinforce their own social position and identity at the expense of other, sometimes even more excluded, minorities. This is particularly true when this is reinforced and fuelled by political support – in Hungary, by governmental support. Other cases mentioned herein describe scenarios in which Romani people feel they are being upstaged by other minorities considered to be more worthy of assistance who then attack the Roma with impunity (in the Czech Republic, Ukrainians, in Spain, Moroccans).

Although the fight against the stereotypical portrayal of minorities is on the agenda of many media professionals and NGOs, the improvement of relations between minorities and the strengthening of solidarity is rarely a theme in such discourses. 

Possible counter-measures, continuation

In the second year of the project, an international workshop will be held in Bulgaria in the summer of 2024 where the partner organisations will analyse the lessons learned and develop a strategy for local action against disinformation.

The Hungarian partner organisation (the Autonomia Foundation) has already not only developed an anti-disinformation training curriculum, but has also conducted several sessions using it, mainly among young people.

The fact that the Autonomia Foundation, as an independent NGO, cannot receive permission from government-controlled educational institutions and their supervisory bodies to organise courses in the schools is an obstacle to their wider dissemination and further training. Therefore, we can only organise such activities in local communities.

We are pleased to see that the Department of Media and Communication at Eötvös Lóránd University is open to this topic, and we will be holding our second academic seminar there on 26 April. Students will also participate in local actions with the local communities in two project locations on 3-6 and 15 May.

The national partners also will undertake local actions based on these cases and on the strategy to be defined during the workshop in Bulgaria.

As an example of such an exercise, we present a tool developed by the Autonomia Foundation to detect and analyse disinformative content. Participants in the sessions will receive news items for analysis. Their task is to find the elements in these news items (click-bait, allusions, etc.) that indicate they contain disinformation content.

Examples

1 According to NASA, life on Earth could soon reach a critical juncture if humanity does not reduce its water consumption to half a litre per person per day.

According to NASA does not mean anything without a reference (link or exact reference to the original publication) of the NASA or other entities. 

Half a litre of water is not enough for a human being to consume daily. The idea is impossible and should be easily recognized as such. 

This news is typical “doomsday” content and does not include exact recommendations on what should be done.

2 The words of Jesus were found on a three-thousand-year-old stone tablet that an Israeli farmer’s daughter recently plowed out of the ground. The tablet says that before the end of the world, the Jews and the Philistines will fight a war and destroy a hospital. It is time to think and give power to those who believe in the coming of the one true God.

The content of this news is hard to believe because it is excessively correlated with the present situation in Israel with regard to the terror attacks of 7 October 2023 and the subsequent reprisals for them. Moreover, mentioning a “hospital” on an ancient tablet is very unlikely, given that hospitals did not exist at that time.

Communications that recommend giving power to just one religious group are in fact exclusionary and not in line with democratic society. This is also a typical purpose of disinformation messages, to promote the disruption of democratic principles.

Of course, the words of Jesus could not have been written on a 3,000-year-old stone tablet, given the date of his birth. Anyone who recognises this logical fallacy usually gets chocolates at the training sessions. 

3 In an interview, the Head of the Institute of Indigenous Hungarians said that the Hungarians are not related to the Finno-Ugrics or the Sumerians, but that it can be clearly shown that Hungarian people descended from Attila the Hun and that we still carry his genes. Being descended from the Lord of the World also involves responsibility. It is not for us to chase after Europe, but for us to take the lead. Those who cannot do so are disowning their ancestors.

In this news, we see once again the message of exclusion, of a certain group of people being superior.

The news is a mixture of what appears to be a scientific background (genetics) and a messianic view of history. It assigns power not to merit or elected right, but to descent.

This news also mobilises, it wants to achieve something in the recipient’s thinking. This is particularly visible when an emotional argument (“disowning one’s ancestors”) is used in the political space.

It may also be suspected that such an institution does not even exist, and it is easy to find out on the internet whether it does or not.

4 A fight broke out in a german bar yesterday, which ended in a stabbing. The stabbing men were all muslim-looking. This is what happens when foreigners invade our countries. That’s all.

This is a typical FB post in Hungary. It should be noted that the language used in media and social media content is also important. A news item that contains many grammatical or stylistic errors can hardly have passed the editorial system expected of a professional media outlet, i.e., no one has checked either its grammar or its veracity. Here both the words “German” and “Muslim” are incorrectly in lower case, and the article ends with an uncharacteristic turn of phrase: That’s all. The difference between news reporting and personal expression is a good illustration of the difference between formal and informal language use. (Note that this is not true of the media in general, but of news, since there are other formats of the media than news: publicity, entertainment.) The text cannot be news because “in a german bar” is not the kind of error that should appear in professional journalism. If such a sloppy, reference-free designation is encountered, then it should be suspect.

This “news” is specifically biased against Muslims, i.e., it implies that all people belonging to this group pose a threat. The term “invade” is also used against migration, as it is a negative, hostile verb that implies the peace of society is being threatened, as in a war. The term “muslim-looking” is also important. The Muslim religion has no “look”, nor does any other. This use of “look” suggests a general exclusion of minorities and of foreigners.

5 A famous Hungarian celebrity has crashed his car. The at-fault driver was drunk and on drugs. All that is known is that he has often appeared at anti-government protests. Of course we can’t say who he was, but we’d be interested in your tips.

This news links social deviance with anti-government opinion. Without anything  being reported, it is capable of creating the impression in the reader’s mind that anti-government people are clearly drinking alcohol and taking drugs when they drive, breaking all social rules. In other words, a political opinion is coupled with a socially generally condemned form of behaviour, and a negative perception is imposed on a politically different opinion.

The most manipulative part of the text is that it leaves room for conjecture, thus allowing an almost infinite number of anti-government actors to be smeared. The commentators who are close to the government will start the uncontrollable spreading of rumours in the social media, not even started by the ‘news’ provider, of anyone whose name is mentioned.

List of the collected cases

The project is funded by the European Union’s Citizens, Equality, Rights and Values (CERV) programme. The European Commission’s support for this project does not imply an endorsement of its content, which reflects the views only of the promoters, and the Commission cannot be held responsible for its content.
Commission is not responsible for any use which may be made of this information.