Highlights from the Political Accountability Foundation (PAF)’s social media monitoring report:

• Facebook (FB) was an important channel for candidates in their communication with voters – its importance during both 2020 Presidential elections was only exacerbated by the pandemic situation which led to the introduction of restrictions on physical campaigning. Some candidates managed to increase their number of followers more than three times during the campaign period.
• The level of activity of media outlets on FB is incomparably higher to that of candidates. This might be a result of the resources available to media, and the wider spectrum of topics media outlets are communicating to users. Out of the total number of posts published on Facebook, only five percent came from candidates, while 60% from mainstream media outlets and 35% from fringe outlets.
• Restrictions imposed on freedom of movement and freedom of assembly due to COVID-19 pandemic during the first election (March 26 – May 10) forced presidential candidates to run their campaign mostly via traditional and social media. With limited access for opposition candidates to public TV and radio, it made social media the main channel for communication with voters for some of the candidates. Despite the gradual lifting of restrictions during the second presidential campaign,social media remained an important communication channel. The average daily number of posts per candidate during the second election was nearly twice as high (7.68) as during the first election (3.83). This could be explained by a much shorter campaign period for the second election (24 vs 45 days) and the general belief among candidates and voters that the second election will actually take place.
• The COVID-19 pandemic dominated the presidential campaign during the first election, leaving limited space for other issues. Very radical drop in interest in COVID-19 issues during the second election (from nearly 18% during the first election to some 1% in all candidates’ posts) could be a response to the government’s messaging that the epidemic was retreating. The effectiveness of this message was also reflected in the high turnout during the second election and a very low interest in postal voting among voters.
• Three candidates, R. Trzaskowski, Sz. Hołownia and K. Bosak, who had the highest number of posts, managed to significantly increase their group of followers on Facebook during the campaign period (even threefold); however, the example of R. Biedroń or the incumbent president, A. Duda, who had a high number of posts but did not manage to increase their number of followers, show that activity on FB is not the only important factor in attracting followers.
• The incumbent president changed his campaign strategy for the second election and undertook more efforts to increase his visibility on FB. This change did not have a significant impact on his “attractiveness” on FB but overall, his limited ability to attract followers on social media proved not to be an obstacle for his victory.
• The number of likes per post was significantly correlated with the increase in the number of followers, increasing the impact on potential voters – R. Trzaskowski, with the highest number of likes per post, 8 333, was also the leader in expanding his group of followers.
• In general, the candidates’ relatively limited ability to reach wider audiences on Facebook, even during the peak of an election campaign, shows that to succeed, they will have to diversify their channels of communication beyond social media to reach other voters. A challenge that has been made even harder because of the COVID-19 pandemic.
• The importance of social media to spread campaign messages to voters seemed to depend on the demographic structure of the target group of voters – the incumbent president, whose voters are predominantly older generations, did not manage to attract new supporters through social media, despite his significant activity on FB.
• No significant presence of hate speech or black PR has been noticed. Findings show that posts that could be labelled as ‘discreditation of political opponents’ were present in less than five percent of total posts, and in less than two percent of posts from candidates. This was especially noticeable during the 2nd round of the second election, where there was a radical decrease in the number of posts with this label. The latter is especially interesting given the competitive nature of any second round of elections – the absence of a rhetoric aimed at shedding a negative light on one’s main opponent could mean that at this last stage of campaigning the two main candidates were mostly focused on increasing their voter base by appealing with their own electoral programmes to those voters who do not usually participate in elections, instead of trying to steal voters from their opponent.
• The tone of candidates’ posts was predominantly neutral, with a low share of negative messages. Especially the two final candidates during the 2nd round of the second election focused more on positive accents. On the other hand, media outlets presented a very limited share of positive messages, having a visibly higher share of negative accents in posts. This trend only increased during the second election.
• Media outlets were very active on Facebook during the entire campaign period, publishing radically more posts than candidates (some media outlets published more posts than all candidates combined). However, the quantity of posts did not reflect their value and attractiveness for FB users, which can be observed by looking at a lower average number of likes per post, compared to candidates.
• ‘Fringe’ outlets, were relatively less interested in election related topics, except during the days immediately preceding the voting day, however, the issue of ‘discreditation of political opponents’ played the most significant role comparing to two remaining groups of accounts – candidates and mainstream media – almost eight percent of posts discussed this issue.
• There were no visible signs of any coordinated campaign of ‘external factors’ during the election campaign. The activity of media having links to foreign powers, like ‘Sputnik’ was rather low, with no significant impact on the campaign.
• Almost one third of the top-liked posts were videos which shows that social media audiences are increasingly likely to engage with audio-visual content, even if consuming such posts takes more time than the consumption of other types of posts.