The Dynamics of Facebook Protest

This is a report on the experiment that Nodus Labs conducted on some of the more active Russian protest Facebook groups formed after the rigged Russian election in 2011. We made two network visualizations for three different protest groups over a period of one month in order to observe their dynamics. We found that the most influential members of these groups were not too politically engaged before the elections and were mainly journalists, students, event organizers, and media workers. We also found that the groups formed around ideological causes (such as “Putin must leave”) stagnated in their development in January 2012, while the groups formed around a call for active participatory actions (“Volunteers for the fair elections”) have grown in size and density considerably, building a very well connected and yet open network that was able to bring many new members together around their cause. We believe that our report will be interesting to those interested in the nature of Facebook protest movements and for those who want to find out how to build self-sustainable self-organized networks that are actively developing and that can bring people together for a cause.


Background Information

Russia has been on the news recently for the protest movement that’s been taking over the country. While there’s been a lot of discontent about the political system during the last years, there was a sort of unspoken agreement between the emerging middle class and the government to stay away from each other. In other words the government (and Vladimir Putin) could do more or less what he wanted, while the nascent middle class could also calmly earn money and build their lives on top of the post-Soviet ruins. Both did quite a good job during all these years avoiding each other, but recently there was a growing realization that Russia could not move forward without resolving some deep structural conflicts, stopping unabridged corruption, reforming its judiciary system, police and the army. However, when the parliamentary elections in December 2011 were rigged in favor of the ruling United Russia party, this was the last straw for many people who have kept quiet for all these years. Suddenly the protest movements started to form and Facebook was one of the platforms that was extensively used in this process. Dozens of different groups popped up, some older ones got a new impulse. “Against Putin”, “For Fair Elections”, “Demos for Fair Elections” and so on. These groups played an active role in organizing the protests that followed and bringing the people together on the streets. It was especially important because for the first time in many years the middle class felt solidarity. Many people found that they were not alone in their discontent, that there were thousands of people just like them who don’t only care about the money and their material well-being. Who also want to live in a free, open, and tolerant Russia that is developing both socially and economically.

The events that followed brought people together for mass protests with more than 100000 participants. Initially dismissing the demonstrations, the government eventually had to give in and introduce some measures that were demanded (such as bringing back the governor elections, loosening up the political party registration process etc.) The main demand to cancel the rigged elections, register banned parties, and hold the new elections in 2012 was not fulfilled, however. Moreover, the coming presidential elections in March 2012 are not considered to be fair by many, as the only candidates that were allowed to participate were literally hand-picked by the Putin’s administration (if not directly then through the complex web of regulations and permissions that makes it nearly impossible without a prior administration approval).


Our Findings

We at Nodus Labs were interested to analyze some of the more active protest groups formed on Facebook in order to understand their structure, detect the most influential communities and members behind them, and observe their dynamics over time. We took two samples: the first one on the 24th of December 2011, just before the second (and most populous) demo, and the second one on the 26th of January 2012, one month after.
The first sample was analyzed in our article on the Russian Protest Movement and you can find more general information about these groups and the explanation of our methodology there. This time, however, we focused on the dynamics of these networks to see what developments they went through during this month. The results were quite surprising.

First of all, all the groups were formed around a certain slogan: “Putin Must Leave”, “For Fair Elections – Volunteers and Activists” and “For Fair Elections – Germany”. While all of these groups were quite active during that month we were interested in the structural changes they underwent during this period. A good indicator for how “healthy” a group can be is not only how often the existing members interact with one another. It’s also the growth rate of the group, the growth rate of the connections between the existing members, active reconfiguration of the existing influential clusters and members, equalization in distribution of power, active inclusion and exclusion of members (rotation), emergence and bridging of structural holes. Networks that are undergoing these changes dynamically will be much more adaptive, innovative and have a higher potential for a collective action (especially if they become more densely connected). See our research on Inclusive Exclusivity for more on this topic.

What we found out was that the groups formed around an ideological slogan (such as “Putin must leave”) and where the most influential members were political activists almost did not change in size, structure or connectivity. While they recruited some new members, their structure remained very stagnant, maintaining the existing power centers and the same degree of connectivity. For example, the number of “orphan” members who don’t know anybody else from the group in the “Putin Must Leave” group did not change and stayed at a very high level of 58%, which stifles information propagation through this network (as we know Facebook favors the news that are “liked” or “shared” by as many as possible of one’s friends, therefore if a piece of news appears in the group and it’s not “liked” by the user’s friends, they are not as likely to see it). Another group formed in Germany to co-organize demo participants did not also really change in size, structure or connectivity.

We also found that the most influential members in these groups were not actively involved into politics or activism before. Most of them were journalists, students, event organizers who just had enough. Only the “Putin Must Leave” group had a strong cluster of pro-Kasparov (opposition leader) followers and a smaller cluster of the activists from Georgia (who probably are interested in changing the political hardline attitude of Russia towards their country and hope that when Putin leaves it will change).

At the same time, the group formed to coordinate volunteer and activist actions during the next elections grew more than 50% in size and more than 120% in connectivity. It became much more interconnected, its density was the highest of all groups, more than 20% of the orphans members were integrated into the network even while it was growing.

There are three conclusions that we can make from these observations.

First, we want to refute the accusations made by the Russian government officials that these groups were formed by politically engaged people, opposition leaders or the US government. They are in fact formed by the normal people who had little to do with politics before. For privacy reasons we do not show the names in our charts, but we carefully checked the profiles of the most influential members and found out they were mainly students, journalists, media workers, and event organizers.

Second is that the Russian protest movement seems to be active not as much for ideological reasons (against Putin) but more for the practical goal of organizing fair elections. This can be clearly seen below on the charts and in our analysis. While all the groups got traction in the beginning, right after the election, the only one that was able to actually develop during this month was the group organizing people for volunteer actions during the next presidential elections. Which is good news actually in that it shows that this protest in Russia is not just about expressing one’s discontent. It is actually about society organizing itself to perform the function that the government utterly failed in: making sure that the elections are fair.

Finally, a general observation that when one forms a group around a certain cause it’s much better to make it something ongoing rather than simply protesting. A cause that can unfold in time and has a positive connotation is able to recruit people for a much longer time than a mere slogan. Therefore, forming a group around the cause of “Promoting fair treatment of animals at Farms” is a much more long-lived enterprise than forming a group that has “Animals are our friends” as its cause. There will be much activity in the communities built around specific actions and not mere slogans. Both may get an audience in the beginning, but a specific cause will bring much more traction.

We will now look into each group in more detail below.


Putin Must Leave Group

For the group “Putin must leave” during the period 26 December 2011 – 3 February 2012 the following changes could be observed.

Against Putin Facebook protest group visualization, December 2011

Against Putin Facebook protest group visualization, December 2011

Against Putin Facebook protest group visualization, January 2012

Against Putin Facebook protest group visualization, January 2012

The structure of the group, the influence of the most prominent clusters (political activists related to Kasparov, journalists and people from the media) , and the most influential individuals remained the same. The group grew from 1809 to 1969 members (about 9%), while the number of connections increased from 5358 to 6259 (15%). The average path (the number of people required to reach from one randomly selected node to another) stayed the same (3.7) and the average degree (average number of contacts within the group each member has) increased to 6.3 by about 5%. Just like last time there are still 58% of “orphans” within the group – the people that don’t have any “friends” within the group, making it very difficult for information to propagate through the whole group and stifling the possibility of a collective action. This is confirmed by another data: the network diameter (or the longest path between members) increased from 10 to 11, meaning that while the new members were added the connectivity of the group was not improved.

Therefore during this month the group was busy acquiring new members (relatively successful in that) but the connectivity of the group at large did not increase that significantly. Also, the structure of the group did not change much. There are the same influential clusters and members as one month before, so there are no newcomers.

Try InfraNodus Text Network Visualization Tool developed by Nodus Labs. You can use it to make sense of disjointed bits and pieces of information, get visual summaries for text documents, and generate insight for your research process:

In order for this group to become more functional it has to work on integrating the “orphans” (members who don’t have any of their Facebook friends in the group) into the already existing clusters and also encouraging them to add their own friends into the group. Also, in order to promote development and innovation there should be more inter-cluster activity between active members and the new influential members should emerge in this group. It is quite evident that the main attraction of the group at the moment is the slogan it proposes, there is otherwise no activity within that would make the members more interconnected or that would change the social constellations established in the past when the group was forming.


Volunteer and Activist Group

For the group Митинг за честные выборы. Акции и Волонтерство (Meeting for fair elections. Actions and volunteers) group on Facebook – period 26 December 2011 to 3 February 2012.

Volunteers and Activists Facebook protest group visualization, December 2011

Volunteers and Activists Facebook protest group visualization, December 2011

Volunteers and Activists Facebook protest group visualization, January 2012

Volunteers and Activists Facebook protest group visualization, January 2012

The first striking difference of this group from the other ones is that it showed the highest rate of growth during this period almost 50% more members and more than 120% new connections between them.

Another quality of this group is that during one months the most influential clusters reconfigured themselves, included the new members and excluded the old ones.

The magenta cluster became the most prominent one and one of the nodes (who is a journalist, writer and publisher working mainly for independent publications in Russia) within this cluster got a lot of influence.
The blue cluster was also reconfigured and one node (also a writer and a researcher in economics) became much more prominent than the others.

Overall, the network got much more interconnected, the existing members made “friends” with the other members of the network, the most influential members also changed. This is a very positive kind of activity for a network that wants to develop and to be open to the new influences. Increasing its internal connectivity, while redistributing the influence within the network.

Also, one month ago this network had 45% orphan members (those who didn’t have any Facebook friends in the network), which was a relatively high number. Now the network has 36% orphans members, which is an important achievement considering that it grew more than 50%. This is another sign that the actual cause that this network proclaims (volunteering for fair election campaign and recruiting election observers and activists) is in demand with the members of the group and that they even connect to each other after joining the network.

It should also be noted that this group has some of the lowest average paths and the highest average degree in the network (that is, the average number of mutual friends who are inside the group) and the longest distance between randomly selected members of the group is the lowest.


For Fair Elections (Germany) Group

For the group, “For Fair Elections in Germany” during the period of 26 December 2011 – 26 January 2012 the following changes could be observed.

The two main nodes that represented the two most influential clusters (magenta and blue) kept their influence in the network. The influence of the most influential representatives within the two other prominent clusters reduced.

For Fair Elections (Germany) Facebook protest group visualization, December 2011

For Fair Elections (Germany) Facebook protest group visualization, December 2011

For Fair Elections (Germany) Facebook protest group visualization, January 2012

For Fair Elections (Germany) Facebook protest group visualization, January 2012

Some active members of the group (that were not the most influential ones before) created new connections within their own clusters and outside them. This allowed them to become more influential in the network. Because of that currently the distribution of connections among members in the group is more equal. There are several equally influential clusters, which contain several equally influential participants. This is a good sign concerning ideological polysingularity of the group (the presence of several simultaneously possible solutions), as there is enough diversity to allow for plurality of solutions yet sufficient number of connections between the different clusters to allow for collaboration.

The total number of members in the group was increased by 5% (to 495) while the number of connections within the group increased by about 20% to 1349 (mainly due to the active members making new connections between each other).

On average every member of the group got about 10% more contacts within the group than before (average degree increased from 4.8 to 5.4). There are also less “orphans” (those who don’t have any “friend” connections within the group). However, there are still 200 of them (so more than 40% of the group don’t know anybody else in the group). This, of course, makes it more difficult for information to propagate through the network and stifles collective action.
Some of the members however created the new connections with representatives of other clusters so that the total modularity (prominence of community structure within the group) decreased and the graph density increased by 10%. That means that the group became slightly more homogeneous while still retaining its heterogeneity.

In general the trend for this group is quite positive because there is an increasing number of interactions and connections between the members, both within and outside of the existing clusters. Another good trend is the increasing equalisation of influence among the group, thanks to the emergence of the new prominent members. Also, the most influential clusters are becoming better connected without any negative effect towards the other clusters which are also growing in size.

Among the negative trends is a low trend of the “orphan” integration. There are still 40% of those in the group who don’t know anybody there. So the activity of the group should be directed towards integrating those orphaned nodes into the already existing clusters or helping them create their own. This way they’d be able to have more influence in the group and also help information spread faster and easier through the whole community.
Also the growth tempo for the group is quite low (5%), so the already existing members could invite their friends to join the group. By the way, if the orphaned nodes invite their friends into the group this will partially resolve the fact that they are removed from the influential clusters.
Finally, the members are mostly the same as 1 month before within the clusters, so more inter-cluster interaction could be recommended to maintain heterogeneity and the influx of new ideas within the group.

On the internet people come and go, but we would like to stay in touch. If you like what you're reading, please, consider connecting to Nodus Labs on Facebook, Twitter and Patreon, so we can inform you about the latest updates and engage in a dialogue.