Georgia’s pro-democracy protests larger than reported

Fringe Facebook pages, pro-Kremlin outlets, and Georgian officials underestimated the size of the demonstrations

(Source: @KaranKanishk/DFRLab via Natia Chkuaseli)

As the largest anti-government demonstrations in years swept through the Georgian capital of Tbilisi, Georgian officials, pro-Kremlin outlets, and fringe Facebook pages pushed unsubstantiated claims about their size in an effort to undermine support.

Georgian politics have been turbulent over the past decade, with Russia’s 2008 invasion of the country acting as an accelerant, and its increasing interference continuing to feed the flames. Russia drives controversial and polarizing narratives in part because it sees instability in its neighboring countries as a way of increasing its relative power and profile. If it can successfully sway an election toward its preferred candidates, or — at minimum — disseminate political chaos, organized resistance will have a hard time forming. Russia’s preferred result from pushing such incendiary narratives is that Georgia will remain too polarized and unstable for Western institutions, such as NATO or the EU, to allow for its accession.

Georgia has seen many protests in 2019 to date, with a particular spate of protests in late June, when Russian Member of Parliament Sergei Gavrilov was invited to speak before the Georgian parliament. The subsequent protests resulted in a relatively harsh police response, which in turn led to more protests. By the end of June, however, the protests had largely dissipated as the government promised to make significant electoral reforms in response.

Protests resumed in November after Georgia’s parliament failed to adopt the reforms promised in June by billionaire oligarch Bidzina Ivanishvili, head of the ruling Georgian Dream Party. Ivanishvili’s promise was an attempt to accede to the demands of thousands of Georgians who took to the streets this past summer in protest of a Russian MP’s visit to Georgian Parliament.

Claims of low turnout and fearmongering that “Soros-funded” activists were plotting a “Rose Revolution 2.0” circulated online. The DFRLab found the claims of low turnout for the largest protest on November 17 to grossly underestimate the size of the crowd. Relatedly, the DFRLab also looked at the traction stories about the protests received in the Russian-language information space, both on media outlets such as RT and on social media platforms such as VKontakte, and found that the engagement was relatively low.

Estimating crowd size

While Reuters, AFP, and Radio Free Europe estimated the crowd size at around 20,000 people, Georgian officials as well as a number of Georgian Facebook pages claimed the turnout was considerably smaller. Chief of Patrol Police Vazha Siradze stated that the Ministry of Interior had determined that 5,500 people attended the protest on November 17 but provided no details as to how the ministry arrived at the number.

To verify the number of protesters present, the DFRLab examined the video footage posted by Mtavari Arkhi, one of the most popular Georgian television stations, which livestreamed the demonstrations.

The live video footage of the November 17 protests streamed by Georgian broadcast outlet Mtavari Arkhi. (Source: Eto Buziashvili/DFRLab via Mtavari Arkhi/MyVideo)

Organizers announced that the protests before parliament would begin at 3:00 p.m. For crowd estimation, the DFRLab used footage from a roughly 15-minute interval around that time.

The DFRLab geolocated the site of the protests to be Rustaveli Avenue in central Tbilisi, right in front parliament.

The area of the protests: the pink pinpoint shows the Parliament of Georgia, from which the video was shot, while blue stripes indicate the building in front of parliament. (Source: @EtoBuziashvili/DFRLab via Mtavari Arkhi/MyVideo)

As shown in the livestream, the crowd in front of parliament was tightly packed. In order not to inflate the size of the crowd, the DFRLab assumed a standing density of two people per square meter.

The diagram illustrates the standing density of two people per square meter. (Source: gkstill.com)

Assuming a minimum density of two people per square meter, the DFRLab used the online crowd estimation tool MapChecking to calculate the crowd size within the area visible in the livestream. MapChecking estimated a crowd of 17,711 people in front of parliament.

Crowd estimation of the geolocated protest. MapChecking suggested that the visible crowd at the beginning of the livestream consisted of at least 17,711 people. (Source: @EtoBuziashvili/DFRLab via mapchecking.com)

Crowd estimation is an inexact science, and this analysis has some caveats. First, crowds are not static, and their approximate density varies over time. Second, it is difficult to determine the precise boundaries of a crowd from photos and video footage.

Where possible, however, the DFRLab has chosen conservative estimates to avoid overcounting. The static density estimate of two people per square meter employed in this case is likely low for this particular crowd: this density typically allows for “free flowing movement” with “no restrictions,” but the crowd seen in Mtavari Arkhi’s livestream footage was relatively closely packed. Furthermore, the DFRLab only assessed the crowd size within the section of Rustaveli Avenue visible in the livestream, but the crowd likely extended beyond those bounds.

While the MapChecking tool does not offer a margin of error for its calculation, assuming a liberal one of 25 percent puts the approximate crowd size between 13,200 and 22,100, which is still well above the figure cited by Georgian officials, anti-opposition Facebook pages, and pro-Kremlin outlets.

Meanwhile, some of the same Georgian Facebook pages that the DFRLab previously reported on during the June pro-democracy protests also downplayed the crowd size of the protests in the days leading up to and after the November 17 protest, which received the highest turnout.

Several Georgian Facebook pages the DFRLab previously investigated for spreading anti-opposition narratives claimed low turnout at the late November protests, including by mocking the crowd sizes (top left) or suggesting that protesters used drugs (top middle). (Source: ჩვენი დროის გმირი/Heroo/archive, top left; Informational Portal • საინფორმაციო პორტალი/archive, top center; მახსოვს/Makhsovs/archive, bottom left; სირცხვილის კორიდორი/archive, bottom center; ქართული აზრი./archive, right)

These Facebook pages also spread unverified photos and videos accusing the demonstrators of using Molotov cocktails and drug use. Some of the photos and videos showed drug paraphernalia, such as hypodermic needles, but provided no information on when and where the photos were taken.

Kremlin outlets mobilize against the protests

Kremlin-owned and pro-Kremlin outlets devoted extensive coverage to the November 17 protests. A Sysomos analysis of the keywords “Тбилиси” (“Tbilisi”) together with “митинг” (“protests”) showed a sharp increase in mentions on November 17, which declined to near zero by November 19.

A Sysomos readout showing the spike in mentions of “Tbilisi” alongside “protests” around November 17. (Source: @EtoBuziashvili/DFRLab via Sysomos)

Kremlin-owned outlet RT claimed the protests suffered from low turnout. In an effort to further delegitimize the demonstration, it referred to the opposition as the “so-called opposition” and claimed it merely wanted to take over the country.

The RT article also claimed that the anti-Russian hysteria and Russophobic rhetoric of protesters will only harm Georgia’s economy as the number of Russian tourists decreases. A BuzzSumo analysis of this particular article showed very little engagement.

A BuzzSumo scan showed that engagement with the RT article about the Georgian protests was low, at 91 total engagements across all platforms BuzzSumo includes in its analysis. (Source: @EtoBuziashvili/DFRLab via BuzzSumo)

Concurrently, pages on Russian social network VKontakte accused “Soros-funded Georgians” of preparing for another revolution with the ultimate goal of returning exiled former president Mikheil Saakashvili to power. For a majority of these VK posts, the engagement was zero or close to zero.

Pages on Russian social network VK claimed that Soros-funded Georgians were planning “the Rose Revolution 2.0” and the return of former President of Georgia Mikheil Saakashvili, which they claim would only provoke war in the region. (Source, left to right, top to bottom: Большая политика/archive; НОД СПб и ЛО (официальная группа)/archive; Журнал “Читатель” — 2.0/archive; Геополитика. Политика. Аналитика. Вооружение./archive; ЛНР и ДНР новороссия это часть России/archive; Новости | ДНР | ЛНР | МИР | NEWS-FRONT/archive)

Conclusion

Both domestic and foreign actors attacked the November 17 protests in a similar manner to the information operations that targeted the protest movement that gripped Georgia in June.

In this case, the primary intent in disseminating the claims appeared to be to undermine the Georgian pro-democracy movement, which has the support of both the United States and the European Union.


Eto Buziashvili is a Research Assistant with the Digital Forensic Research Lab and is based in Georgia.

Follow along for more in-depth analysis from our #DigitalSherlocks.