Golden Dawn and the violence against immigrants in Greece. The analysis of Facebook use and its implications for both sides of the conflict.

My Sentiments Exactly: On Methods of Early Warning, Detection and Monitoring through Online and Social Media

Team Members :

Alexandra Kontou, Jonne-Marie Bouman, Jakub Dutka

1. Introduction

In May 2012, Golden Dawn - an extreme right wing organization - enters the Greek Parliament for the first time with 22 members. The elections resulted in a coalition government headed by Antonis Samaras which includes the parties of New Democracy (ND), Syriza Unionist Social Front (SYRIZA), Panhellenic Socialist Movement (PASOK), Golden Dawn (XA) Independent Greeks (ANEL), Communist Party of Greece (KKE) and Democratic Left (DIMAR). Golden Dawn as mentioned already is an extreme right-wing, fasistic, neo-nazi organization; that is how media, university professors and others characterize it. In May 2012, after the Parliament elections, Golden Dawn first entered parliament, scoring 441,018 votes.

The members of Golden Dawn call themselves ‘nationalists’. They have been accused of extremist activity, exercise of racist violence, even assassination attempts (and the murder of Paulos Fyssas in 2013). There is evidence that Golden Dawn acts violently against immigrants, characterizing them as ‘sub-humans’, and threatening them. It supports the cultural superiority of the Greek nation and the white race. Golden Dawn opposes extreme way migration (while supporting the expulsion of all immigrants and Roma), Marxism as to the internationalist approach, globalization and multiculturalism, showing also antisemitic ideological trends. The party supports the legitimacy of the dictatorial regime and also supports the view that the negative response to the Italian ultimatum demonstrates patriotism and political prowess. It defends the action of the security battalions. Finally, Golden Dawn advocates in favor of Neopaganism and ancient religion. Golden Dawn members have been compared to the Second World War Nazis, and their more physically aggressive (often camp-guarding) party the SS, including arm salutes and holocaust denial.

LA.O.S is the ninth party in the Parliament that got elected in June 2012, which has an extra-parliamentary opposition. LA.O.S are the initials for the Greek ‘‘Popular Orthodox Rally’’. Its purpose is the assumption of power by all legal and democratic means. It moves in the wider right-range, but is not as extreme right winged as Golden Dawn. Basic positions of LA.O.S to address illegal immigration:

i) Is the removal of the Law 3838 of 2010 on Citizenship (current provision for Greek citizenship and political participation of expatriates),

ii) The need for an inventory of illegal immigrants living in the country, considering the needs of the economy and the definition by the parliament of an immigrants’ population limit with ranking of the needs of Greece and assessment of immigrants’ skills.

iii) LA.O.S. proposes repatriation of the remaining immigrants.

Greece, among other European countries has been hit by the so called Eurocrisis that came to surface in 2009. The crisis is a combination of government debt, a banking crisis, and a growth and competiteveness crisis. Greece was hit especially hard because of its main income sectors being shipping, food, tobacco processing, textiles, chemicals, metal products and mining, were severely damaged by the crisis that struck Europe. As the Greek government tried to save these industries by financial injections, the country’s debt increased. The success of Golden Dawn, started in this period of uncertainty; 20% unemployment and economic stagnation. Apart from the aforementioned issues, Greece has the highest rate of illegal immigrants per 1.000 citizens on a European scale. As mentioned earlier, Golden Dawn upholds a strong anti-immigrant, sometimes even characterized as ‘white supremacist- ideology’. The party seems to mostly appeal to unemployed people, those who used to be members of the security forces and people from the criminal underworld.

2. Research questions

The research questions we formulated are meant to investigate the correlation between the on-ground events in Greece and the user engagement on Facebook. We aim to scrutinise the two groups involved in the conflict : the supporters of the racist ideology and the protest groups. By protest groups, for the sake of this research, we mean all groups who actively denounce the rightist movements and acts of violence against immigrants, as well as Non-Governmental Organisations concerned with monitoring the course of the conflict. On top of that, that group also includes the victims and groups interested in protecting the weak from the thugs.

The research questions we intend to answer are the following:

  • How prevalent is the online presence of both the groups in question? The answer to this question would give a clear image of the difference in user engagememt of both groups (pro- and anti-).

  • How do the involved groups use Facebook to their benefit?

  • How do on-ground events correlate with online Facebook posts, comments and likes concerning politics in Greece?

  • Is there a correlation between spikes of online activities of the sides of the conlfict and actual acts of violence?

  • How accurate is Facebook as a tool for monitoring extremist activity in Greece?

3. Methodology

In order to provide a detailed account of the use of Facebook in the struggle between the immigrants in Greece and the nationalistic, xenophobic thugs, we decided to begin our analysis from the grassroots level. In order to gain undistorted insight into what is underway, we set up two research accounts - bogus Facebook profiles appearing to be average Greek citizens. In the process, we made sure for our accounts to reflect typical members of the Greek society who could possibly be interested in the issue. Both the accounts are male, young to middle-aged. For clarity's sake, as well as to keep up the appearances, we agreed for one of the accounts to follow and engage with the anti-immigrant content, while the other one would inquire into anti-racist movements.

The first task was to query keywords relevant to each side of the conflict using Facebook Graph Search, in order to build a list of pertinent groups and pages. The queries were conducted in both English and Greek. Then, in order to broaden the spectrum of inquiry, we resorted to 'Snowballing' - we followed the found pages to further find out other pages they 'liked', which would be centered around the same ideology. Ultimately for that stage, we used Google to query for more related pages. As a result, we built 2 lists of Facebook pages and groups that were used for the research.

We began our analysis using Netvizz- a data collection and extraction application, developed by Bernhard Rieder, that allows researchers to export data in standard file formats from different sections of the Facebook social networking service. The first element we wanted to explore was the 'like network' of each page in question. The 'like network' refers to the pages a particular page 'liked', so, therefore, established a form of connection with. The analysis of page's like network would allow us not only to trace pages that might be centered around the same issues, but also be an indicator of whether the page is active and what its relative status on Facebook is (in other words, whether it is 'important' or not).

Another step we undertook was the gathering of metadata of all the pages in question and analyse it against the available information about the on-ground events in Greece. The metadata was gathered using another Netvizz script, that extracts page posts. We decided to extract data from 6th May 2012 (the day the Golden Dawn was elected to join the Greek parliament) and 15th January 2014 (the day of the research). On top of that, we decided to set Netvizz to gather the maximum allowed amount of data (999 posts per page), to get the most accurate overview of the activity.The obtained data was then converted into one extensive Excel file that allowed us to sort it according to our needs.

The two data sets we received in the step above (one for pages supporting the right-wing ideology and one for the protest groups) was then sorted by 'engagement' - summed up total of the number of 'likes', comments and shares for each post on the page. Subsequently, we decided to visualise the top results in a form of bar graphs (again, one for each group in question), in order to find peaks of activity and tie them to their respective dates. That form of visualisation would allow us to contextualise the peaks by finding the correlation between the online activity and on-ground events in Greece. In order to ensure the graphs' readability, as well as to filter out the less engaged with content, we decided to narrow down the datasets to only the top scoring posts. Because of the vast differences in engagement for each of the groups, as it will be discussed below, we decided to limit the scope of the analysis to posts which scored above 100 and 5 engagement 'points' for the protest groups and the nationalistic groups respectively.

Having made the graphs, we used news sources and The Crisis Maps to search for actual events in Greece and see whether there is any link between them and the Facebook posts. Finally, we decided to conduct brief content analysis to better understand how the engagement is produced on Facebook pages. To achieve that, we followed the top scoring posts to evaluate the type of content they represent.

To further analyse the online activity of the nationalistic groups, we undertook a network analysis of the group interactions, using data scraped by Netvizz and then visualised using Gephi platform. We decided to focus on ΛΑ.Ο.Σ (Με τους Έλληνες για την Ελλάδα) (With Greeks for Greece), which has almost 10 000 members. We were looking for patterns that would allow us to make better sense of how the group functions and what is its structure.

4. Findings

The ‘findings’-section of this paper will give some insight on the results that were found after running through the methods. In Fig. 1 and Fig.2 the like networks per research account, per page are being presented. As one can see, apart from the supporting group having less pages available for research, there are also way less ‘liked pages’ in this network than in the opposing one. The highest number of liked pages in the supporting network was 7, while the highest amount in the opposing network was 201. Apart from these tables, the graphs in Fig. 3 and Fig 4. also show the different amounts of user engagement between the two online sides. While the rightist graph only reaches about 140 in user engagement, the protest graph almost reaches the staggering amount of 1800 in user engagement.

Fig. 1 The like network of the groups supporting nationalist ideology

Fig. 2 The like network of groups opposing the racist violence (top 19 results)

Fig. 3 The activity pattern of the rightist groups

Fig. 4 The activity pattern of the protest groups

The graphs in Fig. 3 and Fig 4. show the dates in the horizantal axis and the amount of user engagement (posts, comments, likes) on the left axis. Combining the two, one could see the amount of user engagement on a specific date. To analyze these graphs, we decided to only take the highest outstanding spikes. For the protest groups, we found 7 spikes worth analyzing and for the rightist groups we found 6 to analyze.

Another difference - apart from the big differences in ‘like-networkedness’ and user engagement - is the decrease of user engagement when it comes to rightist parties and the increase of user engagement in the protest parties. Both can clearly be seen in the graphs in Fig. 3 and Fig. 4. The reason for this difference could be the arrests of active Golden Dawn members or the blocking of Golden Dawn members’ Facebook accounts that started in November 2012.

Looking at the bar chart of the protest groups, we notice that the spikes increase as the dates get closer to the 15th of January 2014. The most engaged with content was photograph posts with satiric and offensive content about Golden Dawn and Greek politicians in general. On 22 October 2013 we observe the first noticable spike and the post related to it is a picture showing a Prime Minister sitting behind his desk while giving orders to three men ‘’slaves’’ that are carrying his desk and below this picture another one of a Leader carrying the desk with three men. This picture was posted by the Facebook group The roof is on fire, with a long comment describing Antonis Samara’s

course in politics since he was 26 years old. On 11 December 2013, the Facebook group We say NO to Golden Dawn, posts a picture of the manger where Jesus was born, with the following joke: A rasist asks the salesman ‘’I want a manger without blacks, arabs and Jews’’ and the salesman aswers ‘’Ok!So you want a manger only with sheeps, cows and donkeys’’. On 31 December 2013, the same page posted a photograph of Antonis Samaras talking to a black girl and the girl is saying to him: ‘’My dad says you are a rasist’’ and Antonis Samaras answers: ‘’Don’t believe every niggah!’’. On 1 January 2014, this group published a photograph of the New Year’s cake of the Antifasist group. Two days later, the Facebook group Europeans against the political system posted an allegorical picture of a tree that somone tried to hedge but the tree is ‘anarchic’ so it can not ‘’limited’’. On 6 January 2014, Europeans against the political system posted a photograph of four underaged Swedish boys who were pissing on a police car. The last most popular post was from the group The roof is on fire in 12 January 2014, where there is a a photograph of a famous TV presenter (Eleni Menegaki) carrying many shopping bags from the most expensive Mall in Athens. The comment below says: ‘’This is how things are. Mr Kontominas makes a huge fortune through illicit loans that at the end are charged to us. With his property, Menegaki buys channels to ensure her communication immunity and fuck 25-year-olds.’’. It does not look like a political comment, but there is a story behind this, which regards the political and financial situation of Greece and how some people are exploiting public wealth thining that the mass is unaware.

To compare spikes in the data that was retreived in Netvizz on both accounts between the 6th of may 2012 and the 15th of January 2014, the Netvizz data was exported to Excel and then turned into graphs. For analyzing the spikes, the map of Athens on Crisis-Scape was used. The Athens-Map in Crisis-Scape only gives a certain range in dates that can be used, which is from November 2009 untill December 2013, which means not every found date could be looked at using Crisis-Scape. For the ones we were unable to find in Crisis Scape we decided to look for Google results by looking for ‘[date] Greece’ on

The first spike you can see in the graph above is related to the 22nd of October 2013. On the crisis-scape map we found there had been three violent attacks throughout Athens in the month October. These violent attacks were performed on immigrants by Golden Dawn members. Something we found on Google for the exact date is the forthcoming of tensions between Roma people and Golden Dawn, after a Roma girl was being kidnapped.

The second spike out of seven was related to the economical crisis that sparked a lot of violence over time. This spike is related to the unemployment rate in Greece reaching a record rate on December 11th in 2013.

The third spike we found is closely related to the Greek government and also to ‘a new start’ as it is about Greek Prime Minister’s New Year’s speech which is titled ‘Things will only get better’ on the 31st of December 2013. As for some people, things had only gotten worse since the parliament had been elected, this title could have offended many civilians. Only one day later - during the new year - most of the posts overall are about ‘Grexit’, which is short for ‘Greek exiting the Eurozone’, as there were many speculations about this happening because of Greece being unable to resolve its economical problems and debts. The first of January 2014 would be the predicted date, but so far, Greece is still a part of the Eurozone.

The fifth spike was noticed on the 3rd of January 2014, the only thing this spike could be related to is an event in Greece on the 2nd of January, when Thessaloniki - the capital of Central Macedonia - became the European Youth Capital 2014.

The 6th of January 2014 provided the 6th spike which is not directly related to violence, but is a big national Christian Orthodox holiday in Greece. The celebrations mark two events in Jesus Christ’s life, according to the christian bible. The first one is the visitation of the three wise men to see the baby Jesus, the second event was Jesus being baptized by St. John. The tradition is mostly celebrated by a priest throwing a religious cross in the sea for the men from villages and cities to dive it up. This year, a woman jumped in to the sea to get the cross, which could be why this is the 6th spike in our graph.

The seventh - and last - spike of this graph is related to the emprisonment of two Golden Dawn MP’s and their leader Nikolaos Michaloliakos. BNP’s (British National Party) Nick Griffin defended the jailed leader of Golden Dawn and was often interviewed about it.

For the graph that correlated with those of pro Golden Dawn, we found 5 spikes which were mostly related to violence on the Crisis Map. These spikes also brought up interesting non-violent results on Google results.

It starts in May 2012 when - according to the Crisis Map - there were 5 violent attacks against terrorists. The first spike is on the 19th of May 2012 which is the date where the old parliament is being dissolved and new elections are called for, which will take place on the 17th of June. The formal decrete to dissolve the parliament is singed by President of the Republic Karolos Papoulias and Prime Minister Panagiotis Pikramennos on the 19th of May.

The second spike for this graph is related to the 2nd of June 2012. In the beginning of June (and the reminder of June as well) there were severe attacks against immigrants in the center of Athens. This spike relates to one of these events.

The third spike relates to an events that occurs two weeks later: the new legislative election. This election took place on Sunday the 17th of June 2012 to elect all 300 members to the Hellenic Parliament after the attempt to form a new government failed in may of that same year.

The fourth spike is related to another event which takes place about fives weeks later, on the 25th of July when Greek Olympic athlete Voula Papachristou gets banned from the Olympics because of a racist Tweet. The athlete tweeted something xenophobic relating African immigrants and showed support to a right wing party (Golden Dawn). She got expelled by the Olympic Committee as her values would not coincide with those of the Olympics.

The fifth spike we see in the second graph is related to the attacks to mosques in Athens city centre in August. One of these attacks took place on the 11th of August 2012 where a mosque was attacked by people using gas cannisters. This same month, many Pakistani shop owners were threatened to quit their jobs and leave the country within one week ‘or else...’. The Pakistani are being severely violently targeted during this month.

The 6th and last spike of this grap is related to the 11th of September 2012, which is related to overall onground events in September. September 2012 is a month in which a Tanzanian Community centre in Athens gets vandalized by about 80 people, the threats to Pakistani people continue and there are about 10 severely violent attacks towards immigrants of any descent.

Fig. 5 The network of ‘Με τους Έλληνες για την Ελλάδα’ (With Greeks for Greece)

As is can be seen above, the structure of interactions within the group is centralised around approximately 3 most active users (the darker the node, the higher the ‘degree’- the number of edges attached to it). It is clear that the activity of the group is centered around those dark nodes. This allows us to conclude that the group relies on the activity of those users.

5. Conclusions

We intended to analyse the conflict between two groups in Greece : the supporters of racist Golden Dawn party and the protest groups who oppose the rightist ideology. The scope of our analysis was the use of Facebook by both the groups in question. We wanted to verify whether social media can be used as a monitoring device for the activity of extremist groups. In order to conduct our study, we gathered data from various Facebook pages and groups either supporting or denouncing Golden Dawn and analysed their like networks, activity through time and the correlation between online activity and current affairs at the time.

We found that the opposing groups have almost opposite activity patterns. While the protest groups seem to be gaining ground online and become steadily more visible and active, the Golden Dawn supporters’ activity dimished shortly after the won election in 2012. Furthermore, the prostesters’ network of inter-liked pages is rather extensive, compared to the one of the Golden Dawn supporters. We believe it is linked to the censorship issues that emerged in November 2012.

Ultimately, we came to conclude that there is no causal chain : online activity -> violence; rather, online activity is simply response to on-ground events. Moreover, as the content analysis proved, the most popular content is meme-like pictures related to current affairs.

6. Discussion

It is evident that much additional work will be required before a complete understanding of the phenomenon occurs. The correlation between the social media activity and on-ground events is not a simple one. We firmly believe that although the social media play an influential role in modern societies, its use among middle-aged to older people is insufficient to speak of its total dominance. In other words, we think that the use of social media among the people directly involved in the conflict might be too infrequent to choose Facebook as a platform for analysis of the struggles.

Another aspect that could be scrutinised further is the network analysis. Because we used

anonymised data provided by Netvizz. In order to get a meaningful insight into the structure of the group, it would be advised to use the licensed version of Netvizz that extracts deanonymised data and conduct the network analysis together with more investigative approach of analysing the most important ‘nodes’ of the network and following their online activity. This way, arguably, it would become possible for researchers to reach previously unaccounted for pages.

Finally, an obstacle we came across during the course of our research was that it was extremely difficult to find Facebook space (pages, groups) for vitims of racist crimes. The question we were not able to reply to was whether these groups were not present on Facebook (at least not as a unified group) or simply created groups and communicated using their native language (therefore we were unable to find them in the Graph Search. Ultimately, it is hoped that this study will stimulate further inquiries in the field.


Samuel Blackstone, "Anti-Immigrant Sentiment In Greece Is Growing Steadily Worse." ,Business Insider, 1 September 2012.

Erik Libellarsi, ‘’Leaked Golden Dawn map reveals possible terror plots against immigrants and communist organizations’’, Kasama Project, 12 December 2013.

Nathalie Saravicas and Charlotte McDonald Gibson, ‘’Just what Greece didn’t need: BNP leader Nick Griffin seeks Golden Dawn alliance’’, The independent, 10 January 2014,

Bonnie Malkin, ‘’Two more Golden Dawn MPs jailed in Greece’’, The Telegraph, 12 January 2014.


Anonymous, ‘’Golden Dawn loses state funding’’, Enet English, 18 December 2013.

Kostas Kallergis, ‘’On the streets of Athens with Golden Dawn’’, The Crisis Republic, 10 March 2013.

CMSNY, ‘’Greece’s ‘Golden Dawn’ and the Anti-Immigrant Platform’’, Center For Migration Syudies, 2 April 2013.

Ali Rahimi, ‘’Hate on the streets’’, Human Rights Watch,10 July 2012.

Nikolas Katerina, ‘’Facebook censors Golden Dawn members’’, Digital Journal, 13 November 2012.

ITN, ‘’Greece's Roma community criticises police for taking mystery girl Maria - video’’, The Guardian, 22 October 2013.

Marcus Walker and Marianna Kakaounakis, ‘’Greece struggles to outlaw its Golden Dawn. Conversative government mounts risky effort to declare a criminal organization’’, The Wall Street Journal, 4 December 2013.

Topic revision: r6 - 20 Jan 2014, ErikBorra
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Foswiki? Send feedback