According to the report of The Center for Content Analysis, Ukrainian consulting company, before the second round of the Ukrainian presidential elections in May, 21st, 2019 there was the great spike of the social network users discussion about Petro Poroshenko and Volodymyr Zelensky.
Analyzing content and profiles of this discussion, analysts found:
– Behavior peculiarities of the Poroshenko’s and Zelensky’s supporters depending on their social network, region, gender and outlook.
– Discussion topology peculiarities, particularly, contacts between the Poroshenko`s and Zelensky’s supporters.
In social networks, we found almost equal share of posts with support of Poroshenko and Zelensky, not likely the voting results where they have got 24,45% and 73,22%, respectively.
We emphasize that this data does not describe the real share of the electoral attitudes but only the activity of supporters in a public expression of views.
The short conclusion is:
– Social networks were not the main contributor in the Zelensky’s victory.
– “Filter bubbles”, typical for the Donald Trump presidential campaign of 2016, were very valuable in this discussion.
– Communicative technologies of 2013-2015 years based on the opinion leaders’ long stories and active discussions now gave way to more horizontal and passive communications. Facebook “periphery” became as important, as the “core” of Facebook audience.
– Zelensky’s campaign office targeted his messages separately for different regions. This was one of the solutions allowed him to win even in Western Ukraine. There were sent more positive messages about Zelensky and less negative messages about Poroshenko.
– Fake news and conspiracy theories were widely used in this campaign. It was very similar to Trump’s 2016 campaign supported by Russian trolls.
– Both candidates’ supporters used much more negative than positive messages. Therefore, posts with the positive messages had in average much more interactions than negative, so their role in the campaign was significant.
– Ideology was not crucial for determining the voter’s choice. Almost all the bearers of liberal values were Poroshenko’s supporters, therefore bearers of social values were split equally between two candidates. The patriotic outlook was the sign of the Poroshenko’s electorate but 25% of patriots supported Zelensky. And vice versa, among the Poroshenko’s supporters we saw people who didn’t support his language policy. Conservative outlook a bit more often take place in the electorate of Poroshenko and the opposite modern outlook – in the electorate of Zelensky.
The methods used by The Center for Content Analysis are as follows:
Analysis period: April 11-15, 2019. On this time between the first and the second round of elections, the discussion was very active because two finalists actively worked on the supporters’ involvement. Poroshenko unexpectedly came to the “Right for the power” TV show and had a heated phone exchange with Zelensky. Discussions continued on the format of the debate, on the provocative billboards with the incumbent president’s symbolic and on the drug test. On that time survey found that 20% of voters hadn’t yet made their choice.
Analysis sample. We analyzed posts, but not the comments or reposts in the six social networks often used by Ukrainians: Facebook, Instagram, Odnoklassniki, Twitter, VKontakte and Youtube. In total, during the period under study, in the social media monitoring system YouScan we found 297,771 posts which contained the words “Зеленський”, “Зе”, “Порошенко” and “Порох”, and had at least one interaction (like, “emotion”, repost/retweet or comment). Among this amount we randomly selected 1000 posts for the analysis. As a result of coding, we received 958 posts recognized as relevant, that are really devoted to candidates. Therefore, the error does not exceed 3%.
Coding was conducted manually by the three trained coders and was held in three stages:
- Coding of the content of posts in the sample. Thus analysts identified the presence of the critics and support of both candidates in the posts and messages about the candidates which were delivered within these critics or support.
- Coding the comments under these posts. Here analysts also determined the presence of critics or support of both candidates.
- Coding of other content in the profiles of the authors of sample posts. Analysts counted posts about elections among the last 30 posts in the public feed of these users. They analysts checked the presence of personal content – life photo or stories, as well as posts showing the worldview of the people. Analysts had three parameters of such the worldview: liberal/social, patriotic/cosmopolitan (or pro-Russian), modern/conservative outlook.
User outlook was judged liberal in a case he supported the tax cut or the tax system reform, enterprise freedom, deregulation, and the market-based utility rates. Social was considered to be the position of users who supported rate decrease, wages and pensions boost, etc.
The pro-Russian outlook was not separated from the cosmopolitan because of the national circumstances and was determined by the support of Russia’s actions in Ukraine, dissatisfaction with the language bill considered in the Ukrainian Parliament on that time, by the critics of Ukrainian army and so on. And vice versa, the patriotic outlook of the user was determined by the support of the language bill, Ukrainian orthodox church independence, Ukrainian soldiers, etc.
Conservative or progressive outlook was identified by the attitude to LGBT, women and national minorities rights, family values and so on.
In addition, the monitoring system automatically determined the number of post’s interactions as well as the user`s gender and region where he lives, as indicated by himself.
As reported, Volodymyr Zelensky has won the presidential elections in Ukraine as he garnered 73.22% of the vote. On the evening of April 22 an estimated 2-3 thousand people rallied in front of the presidential administration in Kyiv to express their support for President Petro Poroshenko, who lost the presidential runoff election.