This report presents our main findings from eleven months of social media monitoring. We identify the main narratives of online discourse and their key drivers, as well as the weaknesses of Libya’s social media landscape — and how to address them.
DRI Libya monitored Libya’s social media landscape from December 2018 to October 2019 to better understand public discourse on Libya’s political transformation. Looking at Facebook, and to a lesser degree Twitter, we wanted to find out how ordinary Libyans discuss politics online, which articles and posts trigger their engagement, and which outlets produce the most content and receive the most attention. Additionally, we wanted to assess whether disinformation was being spread online — and, if so, what type of disinformation it was and who was spreading it.
The topics we scrutinised are related to different actors and factors affecting the country’s political transformation: the constitution, as a constitutional referendum was planned for 2019 but failed to take place; elections, as both parliamentary and presidential elections in some municipalities were held and monitored; the UN Special Mission to Libya (UNSMIL), the Special Representative of the Secretary General (SRSG) Ghassan Salamé, who resigned in March 2020, and his roadmap for Libya’s transition, which was to be followed by all UN member states; and finally, the security situation, which gravely affected any political developments, as the might of decision-making was increasingly withdrawn from political negotiation tables and fought over in military battles.
The following key trends emerged from our work:
DRI worked with different experts to collect and analyse social media data over the past year, ensuring a consistent methodology based on both quantitative and qualitative analysis. BuzzSumo, CrowdTangle and Netvizz were used to run category keyword searches, collect data and qualitatively analyse leading stories on Facebook. DRI used The Twitter Archiving Google Sheet (TAGS) and Pulsar to collect data from Twitter’s application programming interface (API). For more detailed information on our methodology, check out the methodology section of the different reports.
With regards to terminology, please note the following definitions of frequently used terms:
|Narratives||Trending topics, including stories around these topics, which connect, and assign meaning to, particular events|
|Engagement||A measure of how sharable a post is equal to the total number of “likes”, comments and shares|
|Interest||Interest in a topic is measured by the level of engagement on that respective topic|
Given the constant upheaval and turmoil of the previous year in Libya, it is no surprise that the focus of Facebook and Twitter engagement and discussion fluctuated throughout the year. Interest in topics such as security, elections, the constitution and the UN presented clear patterns of engagement that reveal useful insights into subjects that Libyans deem a priority, and those that they do not. While public debates on Facebook and Twitter do not provide a wholly representative sample of the entire Libyan population, with 67% of the 6.5 million total Libyans having a Facebook presence, it is possible to derive useful conclusions. This understanding can then shape future policy proposals by providing guidance on how the Libyan media ecosystem can be strengthened and how to appeal directly to Libyan citizens.
In anticipation of elections, a constitutional referendum and other steps within Libya’s political transition, DRI planned for a monthly monitoring report. With the attack on Tripoli in April 2019, which put a sudden end to any political advances and therefore developments in the areas analysed by DRI’s social media monitoring, DRI decided to bundle the reporting period for several months while continuing to monitor public discourse — now with a particular focus on the “parallel war online” and both mis- and disinformation.
May and June 2019
July to October 2019
There are several key narratives that emerge as particularly resonant in online discourse. The clearest of these is the pre-eminence given to violence. Security was consistently the dominant topic in previous reports, beginning in February, in line with the earliest rumours of an LAAF takeover of an area south of Tripoli and its eventual assault on the capital. This speculative buzz regularly produced some of the most engaged-with weekly and monthly articles, before the dramatic shift seen in April in tandem with the LAAF’s move on Tripoli. This shifted an overwhelming amount of public Facebook activity to security-related discussions, at least in part thanks to the Government of National Accord (GNA) and the LAAF’s use of social media to attempt to shape the narrative around the battle and its outcome.
Airstrikes and terrorism were consistently high-performing topics, along with talk of further military intervention from abroad, particularly Turkey. Much of the content produced by individual rather than group-run Facebook pages reported on military activities through a civilian lens; the majority of security conversations were not about the political or military dynamics of the fighting, but rather a more basic awareness of local violence. This should not come as a surprise, given the much more immediate impact that violent activity can have on someone’s life than high-level diplomatic discussions attempting to solve a civil war that has been raging for almost 10 years. After all this time, these conversations must meet a much higher threshold in order to warrant the attention of the average Libyan citizen. People are far more likely to be invested in following information that can keep them alive.
Saif al-Islam Gaddafi posed another significant draw and a consistently relevant narrative, although a lot of this prominence appeared to be based on a PR campaign involving numerous fake accounts (see previous reports). His emergence as a political force consistently drove engagement throughout the year. In January, Gaddafi and Hassam Tatanaki dominated election-related news and produced the majority of engagements. The Mandela Libya page on Facebook ran a poll on his candidacy and had 71,000 respondents, while Gaddafi-related articles regularly received over 15,000 engagements, in three cases topping 50,000.
In March, content relating to Gaddafi earned eight times as many engagements as every other election-related article combined. As expected, this activity died down during and immediately after the LAAF's Tripoli assault but rebounded in July. Half of the top 10 engaged-with articles related to elections and the constitution were about Gaddafi, ranging from supposed meetings with foreign governments to the specifics of his political future. He was somewhat overshadowed throughout August and September but reappeared in the main headlines in October, when Haftar publicly commented that Gaddafi should be allowed to run in a future presidential election.
An important point to consider with the coverage of Gaddafi is that much of it was driven by a few media outlets. Mandela Libya was crucial to his high levels of engagement early on in 2019 but, as mentioned in previous reports, there are serious concerns over the reliability of data originating from that publication. Many of its page “likes” were considered to be from potentially fake accounts, and the majority of its content appears to be Gaddafi-focused in some way. Moreover, much of the other reporting on Gaddafi from January to March was from Russian state-backed outlets, namely RT and Sputnik.
Following the Tripoli assault, 218tv emerged as a crucial voice in the Libyan media landscape, and it paid Gaddafi considerable attention. This was far from the only coverage he received, as Afrigate News, Libya 24 TV, Libya Akhbar and Almarsad also published articles on his activity and statements. But it does suggest that much of his share of the media attention originates in a relatively slim section of it.
The third key narrative to take away from the previous reporting is the low interest in elections and the constitution. Interest in the constitution began the reporting period with a relatively optimistic outlook, thanks to the announcement of a potential constitutional referendum date by the High National Elections Commission; however, this was the full extent of noteworthy developments relating to the constitution. Consequently, the subsequent reporting demonstrated that, without a steady stream of attention-grabbing news, the constitution quickly fell down the list of topics generating public Facebook conversations.
Following the April conflict in Tripoli, even higher numbers of published articles failed to provoke a higher level of engagement. The same is true of elections; at the start of the year, the Central Committee for Municipal Council Elections (CCMCE) drove a respectable amount of online activity, thanks to the upcoming municipal elections. It shared candidate and voter-registration details, reminders of upcoming deadlines and information on the electoral process.
When rumours of an impending attack on Tripoli began to circulate, they diverted audience attention and engagements for these informative posts began to dry up. After the municipal elections took place in March and April, there was little imperative for this audience to return later in the year without some more substantial progress towards organising a constitutional referendum or plans for a future presidential election. Even accounting for the expected dip in activity around Ramadan in May and Eid in August, the figures achieved by content related to elections and the constitution are feeble.
This is exemplified by one of the few stories relevant to the topic to gain traction: a discussion of returning Libya to its former monarchy-based constitution earned 30,311 engagements on Facebook in June. This is an outlandish proposal, but it shows that only extreme and provocative ideas currently appear to generate interest. As highlighted by the conversations being held around security issues, it is clear that Libyans focus more on present crises in the news than on longer-term political speculation that happens at diplomatic and political levels.
Having initially generated little public discussion on Facebook, the January 2020 Berlin conference triggered a surge in conversations on the social network once it began. It seems that as soon as some Libyans saw evidence of the potential of a high-level political event to effect change, more of them were willing to pay attention to it. This demonstrates that the apparent apathy in attitudes towards elections and the constitution is not insurmountable, but rather that a lack of progress has cemented indifference. Each failure of new initiatives is likely to deepen that indifference.
Finally, the UN Secretary General’s Special Representative, Ghassan Salamé, was an important topic of debate on Facebook. Much criticism, accusations of corruption and claims of bias were directed at him. He has become the focal point of much of the most engaged-with UN-related coverage. While it is to be expected that Salamé, as Special Representative of the UN Special Mission to Libya (UNSMIL), would feature heavily in reporting on the subject, the fact that criticism of him is consistently the primary topic of discussion is important.
Comments from Salamé, or his actions, were consistently the driving force behind coverage and discussion of UNSMIL and the potential of an UN-backed peace plan. His interview with Al Jazeera in March, during which he criticised the Libyan political elite’s corruption, generated a huge amount of reporting and Facebook discourse. This critique became the lens through which all UN-related activity was viewed. For example, the Comprehensive National Conference (CNC) was a popular topic of Facebook debate in March. The reporting on the CNC with the highest level of engagement was a 218tv article that directly referred to Salamé’s comments to Al Jazeera and speculated on their relevance to the conference. As noted in previous reports, engagements produced by Salamé’s comments may continue for weeks, if not months, thanks to the constant reiterations of his comments as part of other Libyan public figures’ rebuttals.
After the conflict over Tripoli broke out, there was much speculation that UNSMIL was about to leave Libya. Many online voices used Salamé’s comments as the basis of their argument either way, regardless of the comment’s initial relevance. After Salamé reported to the UN Security Council in May, voices on both sides published select translations of his report in order to promote their narrative. He was accused of bias by both pro-GNA and pro-LAAF publications and accounts, and his resignation was repeatedly demanded from several sides. He continued to dominate UN-related news through to October.
A key voice that somewhat faded into obscurity once the LAAF's assault on Tripoli began, but one that is nonetheless crucial, is the Central Committee for Municipal Council Elections (CCMCE). From December to May, the CCMCE drove a lot of the constructive and informative election-related conversation on social media. It was responsive and engaged with its community, effectively communicating and receiving messages. It used a network of local branches to give regionally relevant advice and effectively disseminate information. The CCMCE provided a well-organised, informative and reliable communications hub, something that has proved to be somewhat of a rarity on Libyan social media. This made it a particularly effective voice in the early months of 2019 and gives it the potential to either remerge or function as a model for future alternatives for communicating vital yet unexciting information.
Digital media was dominated by two particular outlets: 218tv and Almarsad. Based in Jordan and funded by the Emirati government, 218tv emerged in full force following the LAAF's assault on Tripoli, during which over 50% of social media engagements were attributable to it. It also generated several of the main headlines in May by running inflammatory material accusing Salamé of bias against Haftar, and it was almost entirely responsible for the boost in Gaddafi’s profile in July. After being almost invisible in the analysis until April, 218tv dominated almost every category through to October, rarely ranking outside the top two outlets for a given week or month on any topic. It often produces content with no marked difference to that produced by other outlets, and often in smaller quantities, yet it repeatedly earns higher levels of engagement. This suggests that it is much more effective as a brand than other platforms.
It is worth noting the relatively average performance of 218tv’s social media posts beyond sharing articles. This reflects an ability to generate and distribute content that attracts remarkable attention, but a lack of organic community around the publication. Its reporting is extremely popular, but no more people are turning to 218tv as a hub of discussion than to any other publication.
Almarsad is the other digital media publication that demands further attention. Unlike 218tv, Almarsad has an extremely lively and engaged online community; its high level of Facebook engagements extends far beyond articles. It consistently tops the topic-by-topic leaderboard in the July—October report. Almarsad has been a reliably high achiever throughout the entire period covered by the reports. Its article covering the House of Representatives’ constitutional proposals earned a third of the constitution-related engagements for January. It also acted as a focal point for anti-Salamé sentiment in January, when it published six of the top 10 UN-related articles, all of which were critical of Salamé.
Almarsad has a tendency to publish a high volume of content on a topic in a short space of time, often along with pro-Haftar or pro-LAAF undertones. It was accused of receiving money from the UAE by Sky News correspondent Alex Crawford when she publicly lambasted it for twisting her reporting to fit a pro-Haftar agenda. Given its consistently strong performance, it is a powerful voice for the LAAF and has a proven ability to dictate certain events as newsworthy.
Other media outlets
Beyond 218tv and Almarsad, Libya 24, Afrigate News, Libya Al Ahrar and Eon Libya were the other dependable, albeit second-tier, performers. A key factor to consider is where these outlets are based. As will be discussed later, the Libyan media ecosystem is rife with foreign-backed outlets and potential for misinformation. The administrators of the Facebook pages of these outlets can provide an interesting insight. For example, Afrigate News has administrators in Libya, but also in Tunisia and Egypt. Libya Al Ahrar’s page is administrated from Qatar, despite it claiming to be based in Turkey. This is not to suggest malign activity or intent, but to point out that there is high potential for foreign actors, whether state or otherwise, to have undue influence in the Libyan media.
In terms of individual social media pages, rather than those of larger publications, the landscape is fairly diffuse. HeemaSaad stands out as a page that regularly generates significant engagements and some of the top engaged-with posts, but it cannot compete with the larger publications. The OnlyLibya page often does match the same levels of engagements, but appears to somewhat disproportionately focus on Muammar Gaddafi and Saif al-Islam Gaddafi, suggesting both a lack of diversity in topics covered and the possibility of clear partisan interests behind the page.
Misinformation and disinformation have a constant presence in Libyan Facebook discussions. Combatting them is difficult, and awareness of the problem and options for tackling it are limited for the average Libyan. Previous reports have highlighted the prevalence of fake documentation. In January, a Facebook page representing the Tripoli Protection Force was launched to try to refute statements falsely attributed to it; however, while such false statements received over 30,000 engagements, the posts correcting these statements received only 10,100.
This demonstrates that even a well-organised and logical response to the issue cannot fully counter the damage done by the spread of false information. Most misinformation evades that kind of response. Sometimes this false reporting can have dramatic real-world impact, such as when false crime statistics circulated on Facebook. These statistics suggested that there had been a crime wave in Benghazi, leading to reports emphasising the perceived instability in eastern Libya and growing pressure on the LAAF government. As referred to in previous reports, the Ministry of Interior’s response was ineffective. It often waited until after a story had spread before stepping in to deny its validity. Even after these corrections, media outlets often continued to report on the incidents, negating what little effect the rebuttals may have had.
As alluded to earlier, the Libyan media ecosystem is extremely susceptible to foreign interference. A prominent example of this is the Mandela Libya page on Facebook, which heavily promoted the political prospects of Saif al-Islam Gaddafi. The page, which explicitly compares Gaddafi to Nelson Mandela, was created shortly after the former’s representatives visited Moscow. The page had over 100,000 “likes” within a month of its creation and its posts received disproportionate levels of interaction in comparison with posts on similar pages. Mandela Libya ran a “sponsored” poll that received 71,065 responses, and the results of the poll were picked up by Sky News Arabia, Afrigate News and others. Over the next few months, Mandela Libya was one of the top sources for news about Gaddafi, along with RT and Sputnik.
A Proekt report from September 2019 highlighted Mandela Libya as a likely front of Russian disinformation activity. Proekt cited DRI’s previous reporting that drew attention to the outlet and quoted Aleksandr Prokofyev, a known agent of Russian information operations, claiming to have been in contact with the site’s founder, Abdulmajid Eshoul. Prokofyev denied having any conversations with Eshoul about Mandela Libya; however, the site’s homepage has previously been littered with numerous links to the Fund for the Defence of National Values, an organisation that is known to be part of Russian disinformation operations in Libya and beyond. The head of the fund, Aleksandr Malkevich, has acknowledged that two of his employees were arrested in Tripoli after meeting with Gaddafi.
This threat was demonstrated in October 2019 by Facebook’s removal of Russian accounts exhibiting inauthentic behaviour. This included 14 Facebook accounts, 12 pages, a group and an Instagram account, all inauthentically presenting themselves as Libyan. These pages had been sharing news from Arabic-language Russian state-backed media and commenting on politically contentious topics, often aggravating supporters on all sides of an issue. Facebook noted that these pages shared comments on “Libyan politics, crimes, natural disasters, public health, Turkey’s alleged sponsoring of terrorism in Libya, illegal migration, militia violence”, and aggressively promoted both the legacy of Muammar Gaddafi and the future of Saif al-Islam Gaddafi.
As shown by DRI’s previous reporting, these are topics that often resonate with the Libyan public and are likely to be swept up as part of broader online discussions. This allows these pages to normalise themselves, as Libyans become familiar with their names and branding. Consequently, subsequent attempts to proliferate harmful disinformation or narratives will be more and more effective. This behaviour was linked to Yevgeny Prigozhin, believed to be the founder of the St Petersburg-based Internet Research Agency and coordinator of Russian activity in Africa via his Wagner Group.
Prigozhin has been indicted by the US Department of Justice and sanctioned by the US Treasury for his role in Russian interference in the 2016 US presidential election. Prokofyev and Malkevich, mentioned above, are known associates of his. While it is a positive step for Facebook to be proactively catching and removing these pages, their efforts alone will scarcely be sufficient to negate or counter the cumulative disruptive effects. This single example demonstrates the potential for an external power to embed itself and quickly earn undue influence in the Libyan media ecosystem, as shown by DRI’s January report.
Likewise, the Libyan media ecosystem is too dependent on foreign-based or foreign-funded outlets, regardless of any malign intent. Of the 20 media outlets evaluated in the July—October report, 13 had Facebook pages with administrators located in other countries, including Egypt, Jordan, Qatar, the UAE, the United Kingdom, Turkey, Tunisia, Spain and Germany. Many of them also had administrators with hidden locations. While this does not inherently point to malpractice or malicious influence, it is hardly conducive to transparency. There are valid reasons for much of the Libyan media to be run from abroad, but nevertheless this is ripe for exploitation and exacerbates its weaknesses.
This is evidenced by the LAAF's assault on Tripoli. It was accompanied by a wave of content from 218tv and Almarsad, in effect acting as part of the campaign in the information space. Social media platforms that are nominally used by Libyans are in fact home to users from various Arabic-speaking countries. As noted in previous reports, a significant proportion of pro-LAAF discourse on Twitter originates in Egypt, the UAE and Saudi Arabia, and anti-LAAF commentary often comes from Qatar and Turkey. This muddles the narrative for Libyans and disrupts their ability to confidently navigate the available information and evaluate what is true; it also demonstrates a precedent for information operations as an extension of more traditional methods of interference.
The lack of trustworthy and effective primary sources of official information and media outlets allows this provocation to flourish; the longer this goes unchallenged, the greater foothold these actors will have in the Libyan media landscape, as more and more people begin to rely on them and perceive them as trustworthy. Even if culpable outlets and information are publicly labelled as unreliable, without an effective counterstrategy, trust in the media will become dysfunctional; a perception that all media voices are fickle and self-interested will take root.
Without pre-existing organisations and initiatives proactively filling the gaps with fact-checking and promoting trustworthy outlets and information, the majority of Libyans have very little incentive to tackle those tasks themselves. Such initiatives need to be purpose-built in order to ensure the growth of a productive and viable media landscape.
Aside from external influences impacting the health of this ecosystem, there are pre-existing issues that hamper its effectiveness. For example, women are extremely underrepresented in social media discourse. The majority of comments in previous reporting were made by men, and the women that engaged often faced abuse, highlighting why many chose to stay away. Anecdotal evidence referred to in previous reports suggests that many women choose to engage solely in private and female-only online spaces. There are organisations that provide some of these spaces, such as Project Silphium, but this divide leads to an inherent weakness in both the Libyan media and any analysis of it, as the discussions observed immediately have less potential to be representative.
The secrecy and segregation of women’s participation in online discourse makes them vulnerable to targeted attacks, including from state bodies. Previous reports mentioned an in-person meeting of a group of female Twitter users at a café in Benghazi. The meeting was shut down by the Ministry of Interior, which publicly shamed the attendees and referred to the meeting as immoral and lewd. This was picked up by 218tv, while other pro-LAAF pages declared this a victory for Haftar’s security services and spread fake court documents relating to the incident. The need for female users to separate and distinguish themselves, while often necessary, makes them in some cases even more vulnerable to abuse.
Scenarios like this demonstrate the value not only of effective fact-checking systems and reliable media outlets, but also of more inclusive and constructive spaces for online conversation than are currently available. The idea of an entirely welcoming and positive community is as much a fallacy in Libya as anywhere else in the world, but there is significant room for improvement in order to bring the digital spaces available to Libyans more in line with what is reasonably feasible elsewhere.
The guiding principles of any steps to strengthen the online ecosystem in Libya need to consider the essentially non-existent media freedoms in Libya. Training and support for independent Libyan journalists and media organisations is essential as a long-term solution, but must be considered with caution. Journalism that is independent from the GNA, the LAAF or an external power is in short supply, and an increase in the number of well-trained and capable independent journalists could have extremely beneficial ramifications.
This could take the guise of extremely localised and small-scale reporting, such as on airstrikes or utility supplies, or training and potentially reporting abroad. This could balance the unknown elements behind many of the outlets operating outside Libya, but it presents a danger of being perceived as equivalent to those operations; however, a greater quality and quantity of Libyan-owned media is essential to the future of the overall landscape.
To counter disinformation, support and training provided to fact-checking organisations, both domestically and internationally, could allow false information to be caught much earlier and for corrections to spread further. A communications campaign focused on educating both journalists and the wider population on how to identify and rectify disinformation would allow an awareness of the issue to take root and provide Libyan-driven solutions, rather than an external fix that is only effective as long as it is maintained.
Inspiration can be taken from the CCMCE’s effective dissemination of information through its national and local networks, and its model can be used as a platform from which to develop networks of Libyan journalists and media. Even excluding journalism, it provided an excellent vehicle for spreading essential information. Organisations that verify and register easy-to-access and easy-to-understand records of crime, airstrikes and service outages could be grown out of this model. If trusted and reliable sources for this sort of information can be established, it may help to stem the torrent of speculation that often occurs on Libyan social media pages. Again, a communications campaign that gives the Libyan population the tools to verify such information would be invaluable, especially with regards to the attribution and rumours of violence that often stoke division.
It may also be beneficial to work with social media platform holders to take a more proactive approach to identifying consistent sources of disinformation and take action to inhibit them. Facebook has made positive moves in this regard, but it cannot shoulder the burden alone and is not necessarily inclined to try. Additionally, cooperation with organisations such as Huna Libya, which ran several polls on constitutional awareness among young people that were referred to in previous DRI reports, could offer significant opportunity. It has a popular Facebook page that hosts constructive and well-tempered discussions to which the page admins often contribute. More spaces like these are vital and could easily facilitate communications campaigns regarding political, social, constitutional or electoral issues. They could also promote concepts like disinformation awareness, reliable information on upcoming major events such as the Berlin Conference, or bridge-building narratives that will improve the effectiveness of the Libyan media and the population’s engagement with it.