Parliamentary elections took place in Romania on 6 December 2020. The Bucharest-based Global Focus Center, in cooperation with IRI’s Beacon Project, monitored online and social media to examine the prevalence of toxic narratives in the political discourse in the election period. Using the Pulsar media monitoring tool, researchers followed dozens of outlets as well as social media pages of prominent political players over the course of several weeks. The results, analysis and methodology, can be found in the four reports published below – and on Global Focus’ site.

Context and election results

Prior to the 6 December elections, Romania had a minority government led by the National Liberal Party (acronym: PNL, European affiliation: epp). It was supported on a case-by-case basis by the Save Romania Union (USR, affiliation: Renew/ALDE), the Democratic Union of Hungarians in Romania (UDMR/ RDMSZ, ethnic party, EPP) and the Popular Movement Party (PMP, party of former president Traian Băsescu, EPP). The opposition consisted mainly of the Social Democratic Party (PSD, affiliation: S&D, but also representing elderly, rural and socially conservative voters). The PSD was joined in the opposition by the Pro Romania Party (a maverick splinter of PSD run by former Social Democratic Prime Minister Victor Ponta, seconded by former Liberal Prime Minister Călin Popescu Tăriceanu).

Previous significant elections. The recent local elections had cemented the position of the PNL, which got almost 31% of the vote. The newcomer Alliance for the Unity of Romanians (AUR) got 1%. For the purposes of this report, the 2018 constitutional referendum, which sought to define the family in the Constitution in such a way that would ban homosexual marriage. The referendum initially seemed very popular, with up to 80% of the people claiming they would vote. However, actual turnout was only 21%, failing to reach the legal threshold of 30%.i The surprising result rippled among the conservative and ultra-conservative political communities.

Results. Although the recent local elections had a similar turnout to previous ones (47%, compared to 48% in 2016), turnout in the parliamentary elections was a disappointment (32%, as compared to 40% in 2016). While the impact of the increasing number of COVID cases may play a role, the fact that local elections were largely unaffected by the pandemic suggests that disappointment with national level politicians was more relevant.

i The threshold had been relaxed from 50% in order to account, among other things, for the fact that it may be difficult for Romanian citizens abroad to cast a vote.

Official Results. Data table: Permanent Electoral Authority
Official Results. Data table: Permanent Electoral Authority
PSD obtained a plurality of votes (29%), followed by PNL and an alliance between USR and PLUSi. AUR was the big surprise, obtaining 9% and confounding the majority of opinion polls. PMP and Pro Romania failed to meet the 5% threshold to enter Parliament. 

At time of writing PNL, USR-PLUS and UDMRii are finalising the details of the governing coalition, which will also be supported by the ethnic minoritiesiii.

Other consequences. Prime Minister Ludovic Orban resigned his position, citing the disappointing results, but remained president of PNL and tried to negotiate to be reappointed as prime minister. He failed in this attempt but will likely be the new president of the Chamber of Deputies, probably the most influential parliamentary role. Dan Barna, co-president of USR-PLUS, will be vice-premier. Some voices expect that during the final unification congress between USR and PLUS he will pay the cost for the (perceived) disappointing share of vote. The Social Democrats, though having lost any possibility to govern for now, seem unphased: no resignation is expected.


Main findings
The rise of AUR

The Alliance for the Unity of Romanians was not a main concern in the initial design of the study. It was a known purveyor of toxic discourse but at that moment did not seem either the strongest or the most typicaliv. However, in the initial stages of data gathering we did cast a wide enough net to allow us to watch how it grew on social media and how its success was perceived.

AUR is a new nationalist right-wing party that fuses a segment of the unionist movement (advocating unification with Moldova) with a group that denies and relativises the crimes of the interwar fascists (National Christian Party, Iron Guard?). The leader of this latter group was also a member of the Coalition for Family, which supported the referendum in 2018. According to CURS exit polls its voters tend to be younger (but not necessarily very young), less educated and live in rural and smaller urban communities. It also seems to have attracted people in communities that are farther from large cities and have a middling (but not low) development level.

Research found that its communicators were able to punch above their weight in social media communications. According to the visibility score calculated by Pulsar’s proprietary algorithm from among a broad set of public political communicators’ and major parties’ accounts, George Simion (AUR co-president) was more present in the Facebookvi debate than the official account of PNL or the account of the most visible communicator of the party, Rareș Bogdan. If we count the more transparently definedvii number of impressions the contrast is even starker, with George Simion simply dwarfing all other competitors.

“Visibility” and number of impressions for party social media communicators between Nov. 23rd  and Dec 9th. Top 15 communicators in our sample. Data by Pulsar.
“Visibility” and number of impressions for party social media communicators between Nov. 23rd
and Dec 9th. Top 15 communicators in our sample. Data by Pulsar.
Media interest for AUR exploded near the elections. Data and graph: Newswhip.
Media interest for AUR exploded near the elections. Data and graph: Newswhip.

 However, the social media performance was largely un-noticed outside its “bubble” and the visibility of AUR in the traditional media only increased when elections neared, as Newswhip data shows.

Once they got AUR on their radar, media outlets tended to be critical, extensively presenting the extremist discourse of party members: ranging from the sexist and racist declarations of regional AUR leader Sorin Lavric, to the homophobic remarks of AUR co-chair Claudiu Târziu, former co-leader of the anti-LGBT “Coalition for Family”.




The Church and the Plague

Church involvement in Romanian politics tends to always be a delicate issue. The state subsidizes all the recognised religions in return expecting a degree of non-intervention or even support for state policies (or, occasionally, for the candidates of the party in power). The Orthodox Church, the formal confession of over 85% of Romanians, also expects a certain degree of support from the state, at central or local level, in affirming its role in societyvviii. This relation was shaken by the 2018 constitutional referendum to define the family, a referendum which did not pass. Those who most supported the referendum felt betrayed by the parties (some of which had pledged support in writing), by the media (which they expected to describe the referendum as a choice between ‘normal’ and ‘abnormal’) and even by society at large. Ever since, the Church has been more tolerant of its fringe elements.

The gap widened during the pandemic. No less than three major pilgrimagesix faced some kind of restrictions, either implemented or just proposed, going against tradition and threatening loss of revenue. The anti-science groups within the Church, now including some of the hierarchy, were given unusually free reign. The last of the three pilgrimages was on St Andrew’s (Nov 30th), just a week before the elections.

The political reaction was prudent. Out of 3,732 posts in the party communicators social sample, only 17 mention the Church, with another two mentioning Archbishop Teodosie (the archbishop that hosts the pilgrimage to St Andrew’s cave). Interest in the discussion, as measured by Pulsar’s visibility metric, peaked three days before St Andrew’s Day when Archbishop Teodosie called on all priests in the archbishopric to join the pilgrimage and issued them written affidavits that they would be travelling “for work purposes”, so that they may seem to comply with existing restrictions. Theodor Paleologu (formerly PMP) issued a rebuttal that garnered attention (1.7 thousand reactions as defined by Facebook). One day before St Andrew’s Day, Eugen Tomac (PMP) issued a (somewhat ambiguous) support statement that gathered 18 thousand reactions. On St. Andrew’s proper, Claudiu Târziu (AUR) published a video of him partaking in the pilgrimage, with only ~400 reactions. From then on the debate lingers.

​  Visibility, as calculated by Pulsar for posts mentioning the church or archbishop Teodosie among our sample of ~100 party communicators on Facebook. Bubbles represent the most popular posts, with those of T. Paleologu and E. Tomac clearly visible on the 27th and 29th,  ​
Visibility, as calculated by Pulsar for posts mentioning the church or archbishop Teodosie among our sample of ~100 party communicators on Facebook. Bubbles represent the most popular posts, with those of T. Paleologu and E. Tomac clearly visible on the 27th and 29th,

We can see how the larger parties, PSD, PNL and USR-PLUS avoid the topic. AUR tries to piggyback on it, but with limited success.

It is interesting that very little disinformation or even counter disinformation is present in the debate. Politicians prefer to make references to events and garnish them with comments that can be prudent and ambiguous to the point of self-contradiction. For example, some of those who attend or support pilgrimages will diminish the importance of the pandemic. But politicians are more cautious. Theodor Paleologu, who dared criticise Archbishop Teodosie, does so nevertheless as a good son of the Church and makes sure to distinguish between the archbishopric and the Synod of the whole Church. Eugen Tomac supports the pilgrimages, but in such terms that he becomes self-contradictory: Romanians should be allowed to visit their Church whenever they want, he says, but respecting all the necessary prevention rules. Finally, Even Claudiu Târziu of AUR does not stand against science and reason but simply claims he attended the pilgrimage and speaks to why it is important.

The media follow a similar pattern up to a point. We can see a new peak of the debate on 7 December. These are not, however, articles mentioning the church extensively, but rather descriptive articles regarding AUR members (and, especially, the newly elected MP Diana Șoșoacă, who is also the lawyer of the archbishopric ruled by Teodosie).


Not so much disinformation, but preposterousness

There was little outright disinformation. Rather, the political discourse was full of preposterous claims showing how bad ‘the others’ (i.e. political competitors) were and how awful things would be if ‘the others’ got to govern.

Examples of such discourse are presented in previous interim reports. In many instances, politicians make sure that their critiques of the present situation are at least somewhat plausible. They also make sure their grand pronouncements could only be potentially tested in the future (or on counterfactual timelines). For illustration, Marcel Ciolacu, PSD president, claimed that too few tests were being made (possibly true) but also that the battle against COVID was altogether “lost”. Victor Ponta, of Pro România, claimed that closing the schools would generate an educational crisis lasting a generation. Prime Minister Ludovic Orban engaged in some counterfactual history and could not “even fathom” the [dire] situation of Romania if he had not taken leadership of the country.

Such discourse is difficult to assess in terms of toxicity. On the one hand it is clear that it increases conflict and polarization within society. On the other hand, it can be argued that a strong debate, with tolerance for grand accusations, can work as an antidote for complacency and corruption. At any rate, these are liberties commonly allowed to Romanian politicians and (too) commonly used by the same.

Disinformation was not entirely absent though. Diana Șoșoacă, an AUR candidate but speaking as a lawyer of the Archbishopric of Tomis in this instance, famously claimed that “there is no pandemic”. But even she obliquely referred to the government’s failure to declare a formal “state of epidemic”i. And, at any rate, she was not a main communicator of AUR (and did not make our top 15 communicators list). The party officially used tamer narratives, like the theory that closing local farmers’ markets was intended to boost (foreign owned) supermarket chains.



The research tested five areas of possible toxic discourse against three samples extracted by Pulsar. One sample is formed by the general mediax filtered down by specific keywords related to the five narratives and areas to be analysed. Another is the discourse of the official parties’ Facebook pages. The third is the discourse of the main party communicators on Facebook -people who were thought to be potentially freer to express divisive viewsxi. The samples used for this report cover the period between 23 November and 9 Decemberxii.

It is on these samples that we tested toxic narratives in five areas, identified by looking at the governing programmes of the main parties and taking into account the political discussion. These areas are as follows:

  • The COVID-19 Public Health Crisis
  • The Economic Recovery (health and issues of economic recovery dominate the public’s agenda)
  • Measures Concerning Education (the public was concerned about whether or not schools can stay open)
  • Corruption and the Rule of Law (this is a typical, and divisive issue during most campaigns)
  • (Expected) Election Fraud Allegations

Three layers of filtering were applied, in two stages.

  • A string of keywords tested whether a certain entry qualifies as a politically relevant development
  • Another string tested whether they fit in one of the areas of political discourse
  • A third string tested whether toxic words appear. For example, “covrig” (pretzel) does not signal anything on its own. But in the context related to the pandemic, it is mentioned as a sarcastic reference to the disease, given the phonetic similarity (“COVID”).

i Some of these keywords were applied from the gathering stage, but others were applied in later filtering stages. This was done on an ad-hoc basis.

Visibility as calculated by Pulsar for articles mentioning the church or Archbishop Teodosie among our media outlets. Pulsar media sample is not comprehensive, so results must be considered indicative. Bubbles represent the most popular posts.
Visibility as calculated by Pulsar for articles mentioning the church or Archbishop Teodosie among our media outlets. Pulsar media sample is not comprehensive, so results must be considered indicative. Bubbles represent the most popular posts.