In Scrooge’s boots: Lessons learned on disinformation from the 2019 European elections

There has been much heated discussion on the possible influence of disinformation campaigns on the 2019 European elections—including those campaigns launched by outside actors (namely the Russian Federation). This is not surprising considering previous election experiences not only in Western states, but globally. As far as we know, the 2019 European elections were fortunately not targeted by a large and coordinated disinformation campaign. Given the significant attention paid to the 2019 European elections by the public, researchers and policymakers, they present an interesting case study that might help us to learn not only how to tackle the issue of disinformation, but also how to understand and analyse it in the future.


Introduction
In A Christmas Carol, the famous tale by Charles Dickens, the main character, Ebenezer Scrooge, encounters three ghosts on Christmas Eve who provide him with an opportunity to reflect on his past and present life and the prospects for his future. Due to this confrontation, Scrooge re-evaluates his attitudes and becomes a better person. This article argues that at the present moment we-like Scrooge-have an opportunity to reflect on the role of disinformation prior to the 2019 European elections. This reflection might allow us to be better prepared not only for upcoming elections, but also for further debates on the phenomenon of disinformation itself.
As in Dickens's tale, the article first deals with the past and shows that disinformation is a threat to electoral processes worldwide. It then discusses the situation in the run up to the 2019 European elections and demonstrates that the responsible authorities paid attention to the possible influence of disinformation. However it also points out the shortcomings of the measures applied. The last section suggests several recommendations for further debate based on the presentation of four studies on disinformation in the 2019 European elections conducted by the cybersecurity company Safe Guard, the civic movement Avaaz, the Oxford Internet Institute and the think tank Globsec.

The ghost of elections past
The influence of disinformation is hardly a new phenomenon. The current boom in social networks has gone hand in hand with the decline of traditional media, which has led to problems of societal misinformation, making electoral interference an even bigger challenge (Wardle and Derakhshan 2017). Simultaneously, the vulnerability of our democratic institutions to disinformation campaigns has been recognised by various malicious actors and weaponised to serve their ends (see, for example, Giles 2016).
The significant use of disinformation campaigns could be seen as early as the Russian aggression against Ukraine in 2014. The most noteworthy example is the array of fabricated stories surrounding the tragedy of Flight MH17. However, the first time that there was a discussion about disinformation campaigns in terms of them affecting the popular vote in Western states was during the referendum concerning the EU-Ukraine Association Agreement in 2016. In the year prior to the Brexit referendum there were proven cases of disinformation campaigns targeting voters. The US presidential election also took place in 2016-this was significant not only due to the widespread disinformation in the information space prior to the event, but also because it served as a benchmark for the debate on this phenomenon. Similar trends were also seen prior to the French presidential election and the referendum on Catalonian independence in 2017. The above-mentioned cases are well known, however it should not be forgotten that the challenge of electoral disinformation campaigns has also been present in smaller European countries such as Sweden and Czechia. Furthermore, this problem does not only concern democracies in Western Europe, but has global implications as well (Jackson 2018).
Researchers are still struggling to evaluate the impact of disinformation on election outcomes (for more on the current state of research see Tucker et al. 2018). However the existence of disinformation campaigns that aim to disrupt the electoral process is a proven fact that needs to be addressed.

The ghost of elections present
Given what has been described in the previous section, it is not surprising that some were anxiously anticipating foreign meddling during the 2019 European elections. This feeling of anxiety was not limited to the EU bubble, but resonated among the general population as well. According to a Eurobarometer survey, 73% of respondents claimed that they were concerned about the impact of online disinformation or misinformation in the lead up to the election (European Commission 2018c).
European Commissioner for Security Julian King acknowledged this fear, stating that '[The European elections] present a tempting target for malicious actors' (Cerulus 2019). As a reaction to these concerns, the European Commission adopted the Action Plan Against Disinformation in November 2018. This plan focuses on the measures deemed necessary to minimise the effects of disinformation in the 2019 elections (European Commission 2018a). One such measure was the strengthening of the East StratCom Task Force, 1 which went as far as dedicating an entire section of its website to the European elections. Another measure in the Action Plan was the establishment of Rapid Alert System mechanisms that would enable the exchange of information between member states about problematic cases, and in so doing would help to uncover potential disinformation campaigns.
Significant attention was also paid to the role of technology companies, which were asked to uphold the self-regulatory EU Code of Practice on Disinformation introduced in September 2018 (European Commission 2018b). Before the elections, Facebook, Google, Twitter, Microsoft and other companies stated that they would uphold the Code. In addition, Facebook-which has been sharply criticised due to the spread of disinformation on its platform-adopted its own policies in relation to this topic (Woodford 2019).
Despite the major concerns, the situation prior to the 2019 European elections was calmer than anticipated. This evaluation was shared by the European Commission as well as the majority of independent researchers, whose findings will be introduced in the next section. The Commission was quick to claim a victory, and attributed the lack of meddling in the elections to the measures taken (European Commission 2019).
However there is still much to be done on the EU level. For example the Rapid Alert System has not worked as planned; there was no clear understanding of how it was to be used and so it has become a store of unrelated information rather than an effective communication tool between member states (Appuzo 2019). As this illustrates, the further development of existing tools is still needed. Support for independent research should also be provided. The EU should not only offer material support but provide a networking platform that connects relevant actors from the member states so as to facilitate the exchange of ideas and experiences. The European Parliament could play an important role in this regard and advocate for policies tackling disinformation, as well as bringing more attention to this issue.

How to be prepared for the ghosts of elections yet to come?
The 2019 European elections were closely watched, not only by EU institutions, as illustrated in the previous section, but also by independent researchers. A description of their approaches allows us to evaluate the current state of the debate and existing studies and identify their shortcomings.
One of the studies that made headlines before the elections was published by US cybersecurity company SafeGuard Cyber. The substantial amount of attention paid to this study is not surprising considering that it claimed that 'half of the citizens of EU Member states have come into contact with disinformation from Russian sources' (Boffey 2019). This claim was based on an investigation of more than 6,000 bots and semi-automated accounts on social media that allegedly have Russian connections. The estimated number of views was derived from a combination of data from Twitter, Facebook and YouTube. Even though this research was widely cited by the media, there are certain problems with its conclusions. Principally the company did not present the list of bots and accounts that were analysed, nor was it willing to explain specifically how these were able to be linked to Russia. Due to the fact that attribution remains one of the most problematic facets in research on disinformation, in combination with the fact that number of views does not equate with impact, it is necessary to approach this study with some caution. More generally, a cautious approach should be taken by journalists when citing various research papers and experts. As with any other topic, in the case of disinformation the hunt for sensational headlines and easy answers can obscure the whole debate.
Another group interested in these issues was activist community Avaaz, in cooperation with the Institute for Strategic Dialogue (Avaaz and Institute for Strategic Dialogue 2019). Its researchers focused on Facebook use in Germany, the UK, France, Italy, Poland and Spain in the three-month period before the elections. The main focus of the study was the networks of far-right political groups which were consistently used to spread disinformation. The researchers not only aimed to analyse this space, but also took action and reported these pages and profiles for violating the Facebook rules of conduct (Avaaz 2019). The involvement of activists (such as Avaaz) or private companies (such SafeGuard Cyber) in the debate about disinformation is double-edged. On the one hand the variety of actors may bring different perspectives, thereby offering the possibility of innovative solutions. On the other hand it creates a situation in which it is no longer clear what the real goals are of the actors involved in the debate. This is particularly problematic concerning the impact that the various studies might have on public discourse and policymaking. More scrutiny surrounding the debate about disinformation and its actors is needed (for an overview of this debate as it applies to Czechia, see Daniel and Eberle 2018).
Researchers from the Oxford Internet Institute focused on political news related to the elections in the UK, France, Germany, Italy, Spain, Poland and Sweden on the social media platforms of Facebook and Twitter in the month prior to the ballot. They specifically looked at 'junk news', which they defined as 'ideologically extreme, misleading, and factually incorrect information' (Marchal et al. 2019, 1). An important shift in this study is that more attention was paid to the phenomenon itself, rather than its (alleged) authors. This shift is a very important step since it takes the debate further than the narrow perspective focused solely on Russian disinformation campaigns which has dominated in the past. While it is certainly vital to remain aware of the long-term efforts to inject Russian narratives into the EU's information space (as remains the focus of Kalenský 2019) through the intricate use of elaborate networks of fake accounts on various platforms (as proven, for example, by DFRLab 2019), this debate should be tempered and more fact oriented, since giving too much credit to the Russian disinformation effort may in fact make Russia look stronger than it actually is (as highlighted by Győri and Krekó 2019).
Some of the most recent research has been conducted by think tank Globsec. It focuses on Facebook pages known to be spreading disinformation in the Visegrád Four countries in the run up to the elections and aims to describe their reporting of the event. 2 Interestingly the pages analysed did not pay significant attention to the event, which reflects the general disinterest in the elections in this region. The notable exceptions were Facebook pages that had previously already advocated a particular political position (Sawaris et al. 2019). This research represents an interesting approach since it allows us to go beyond the narrow perspective of false stories and to focus on other less visible but more widespread manipulations such as the promotion or silencing of certain narratives. This more holistic approach might help to connect the debate about disinformation with other established academic disciplines (such as media studies) and to place individual disinformation stories in the context of the whole information space.

Conclusion
There was a significant sense of anxiety in anticipation of the 2019 European elections due to the possible influence of disinformation. Having looked back at the experiences of the Brexit referendum, the 2016 US presidential election and several other elections, these concerns were understandable. However, even though the campaign was not disinformation free, it is not possible to speak about a coordinated campaign focused solely on this event. Despite this, the 2019 European elections still represent a great opportunity to evaluate the current debate about disinformation.
The very fact that the EU institutions acknowledged this threat and tried to tackle it should be appreciated. However it is important to bear in mind that these elections were only the first step and others must follow. European institutions should not only improve the functioning of the existing institutions tasked to tackle disinformation, but should also focus on civil society and provide it with the needed resources and framework for exchanging experiences. The European Parliament could play an important role in advocating such steps.
The European elections were closely observed by both state bodies and independent researchers. It is plausible that the debate on this topic could be more inclusive and involve various actors (universities, activist movements and private companies) who might bring fresh perspectives. However this also constitutes a new challenge, since these actors are having an increasing impact on the public debate and policymaking. Therefore a more careful evaluation of research standards and closer scrutiny of their agendas and identities are very much needed. More attention is being paid to the role of domestic actors (such as right-wing extremists) and to the overall context of the information space, which suggests that the debate is becoming more complex and is no longer limited solely to disinformation campaigns originating in Russia. Even though Russian influence campaigns should not be forgotten, this opening up of the debate represents a great opportunity to further extend the scope of research and link it to other, already established social science fields.
Such an opening up is crucial, since the problem of disinformation is a long-term one that will come in many shapes and forms. Therefore it is necessary to be flexible and open to new ideas and approaches. The tackling of this challenge remains vital to preserving the credibility of the foundation of democratic polities-the electoral process. The 2019 European elections should be remembered as one of the lessons that can teach us which threats to expect and how to successfully overcome them.