content analysis Facebook Politics Research Social media

Biden’s organic Facebook campaign is behind Clinton’s, while Trump is surpassing his 2016 performance

Analysing 4450 posts from equal campaign periods before the 2016 and 2020 Presidential Elections, the data shows that Biden is trailing Clinton’s performance, while in contrast, Trump is exceeding his 2016 performance.

Analysing 4450 posts from equal campaign periods before the 2016 and 2020 Presidential Elections, the data shows that Biden is trailing Clinton’s performance, while in contrast, Trump is exceeding his 2016 performance.

Facebook is important to party campaigns, with the 2020 Presidential Elections no exception. Facebook was the battleground that showed Trump’s rise in 2016, and it is the place where we can once again see the President successfully communicate to millions of voters.

In my research of British political parties use of Facebook, I have seen the use of Facebook fit ever more closely into the heart of modern political campaigning, with uses of targeted advertising and novel organic campaigning continuing to rise. Much of the journalistic focus seen recently is upon targeted advertising, however key studies have asserted the importance of organic peer-to-peer campaigning. It is this organic battle that Donald Trump won so clearly in 2016, and data now shows the same situation in 2020.

Donald Trump and Joe Biden’s official Facebook pages’ posts are examined. It is important to note that the pages studied for this analysis only reflect a portion of party campaigns on Facebook. This is because analysis does not include the targeted advertising battle currently underway, nor the hundreds of other Facebook pages the parties use to campaign. However, because these two leader pages feature by far the largest number of (like) followers (a combined 32 million), these pages are vital to their campaigns as they are the core of how messages can be spread organically. As such they provide a clear insight into the overall health of the two parties campaigns, allowing us to appreciate follower enthusiasm and the wider reach the leaders have achieved.

The pages studied in this report are but the tip of the iceberg of Facebook political page’s and content interacted with. These leader/candidate pages are only a part of two political parties campaigns’ on the platform. However, these pages are certainly central to their campaigns, and as such provide a clear insight into the overall health of the two parties campaigns.

Table 1 below outlines the pages examined, the time periods used and the number of posts extracted and analysed. Crowdtangle – Facebook’s excellent research tool, was used to gather all the data seen. Given the different posting frequencies of the pages averages are best used, however total engagement is also presented for context and to appreciate the bigger picture.

Number of postsStartEndElection dayDays analysedDays to election day from start dateDays to election day from end date
Donald Trump 201689601/07/201609/10/201608/11/201610013030
Donald Trump 2020201626/06/202004/10/202003/11/202010013030
Hillary Clinton 201681401/07/201609/10/201608/11/201610013030
Joe Biden 202072426/06/202004/10/202003/11/202010013030
Table 1. Pages analysed, dates data examined across, number of posts examined and days until election day

Trump posts in the highest frequency, with this a recent trend for the 2020 election, the other pages post at similar rates.

Average engagement: Biden is stumbling but is being held up by a small but very engaged activist core. Trump soldiers on.

Graph 1. Average engagement with the different pages across the studied periods

Average engagement has grown for Trump across the two studied campaign periods. This has centrally been through Likes and Comments, with Shares on average decreasing. Although many people, more than ever, are engaging with Trump’s content. When it comes to Shares, the Trump campaign has actually been going backwards since the 2016 campaign. This decline in shares is also seen from the Democratic nominee Joe Biden. The decline in shares shows the two politicians have follower-ships that although still likely to engage via other forms, are less keen to take ownership of content and spread it directly to their own networks through sharing. This is a direct threat against the abilities of messages to break outside of echo chambers. It is an issue the Trump campaign should be paying close attention too, as their engagement (although large) may be becoming constrained to a group of the faithful.

In contrast to the Trump campaign (which although mixed) shows healthy qualities; across all the engagement forms except Comments, Biden has failed to achieve Clinton’s levels of 2016 engagement. Biden is behind in Likes and Shares, with Likes the engagement area with the biggest fall. This decline shows that in core forms of engagement, the democratic nominee has gone backwards since 2016. However, all is not lost as Biden has a small but enthusiastic core of online support.

Graph 2. Number of (like) followers of the pages (data estimates via Wayback machine)

Clinton during the 2016 campaign had 5.2 million followers whereas today Biden has 2.9 million followers. Graph 2 shows the followers of the pages across the two studied campaign periods. Biden has a small active core of virtual members who engage with his content, with the small followership Biden has achieved meaning that messages are more likely to enter socio-demographic echo-chambers as the same users will be sharing content multiple times. The Democrats acted far too late in developing Biden’s page. In contrast, Trump has a much larger audience of 28 million.

This activity from followers is important for message dissemination, as Facebook is a powerful campaign tool that allows for everyday folk to take political content and use it to campaign to their friends. People today are delivering leaflet style content to their friends on behalf of parties, the doorstep is now digital. This is a powerful capacity parties should want to utilise. Given this system, there is a clear benefit from having a broader online support base as you will be reaching more doorsteps.

Although the Democrats appeared to have acted far too late to develop Biden’s page, a metric I have called VMAS – “virtual member activity score” shows a more positive picture for the Democrat candidate. This measure takes the sum engagement and divides it by the number of followers, thus showing the number of engagements a page has achieved per follower. Graph 3 shows the average number of likes, shares and shares achieved per follower across the two campaign periods. The data shows that Biden has a small but active online support base who like, comment and share at a similar rate to Trump’s followers, but like and comment at higher rates than Clinton achieved. However, in shares we have seen Biden’s followers engage at lower rates than Clinton achieved. Although Biden’s followers are small in number they are clearly very active, but because they are less likely to share, content is not spreading as effectively.

The data shows that Trump not only has a larger support base, but one that is nearly as engaged as Biden’s much smaller support base. The fact that Trump’s much larger support base is nearly as engaged as Biden’s in terms of liking and sharing his posts, means the Trump campaign has a far stronger ability to spread its message more widely across Facebook. Given the importance of social transmission of political messaging, if these messages are heading to electorally important locations, then Trump has the greater advantage. This is worrying for the Democrats as in 2016 they were level with Trump but have since fallen behind.

Graph 3. Follower engagement activity — Virtual Member Activity Score

Sum engagement – Total Trump domination

The picture changes when you sum the engagement, in Graph 4 you can see the total domination Trump has over the other pages. This is interesting as although Trump beat Clinton on the platform soundly across 2016, his lead was nowhere near what it is now versus Biden.

Graph 4. Total engagement with the different pages across the studied periods

Reaction sentiment – A stable set of affairs

The evidence from Crowdtangle also suggests that the large increase in total engagement with Trump’s posts isn’t down to people who necessarily dislike him. While there has been an increase in comments (which may be negative), when looking at the type of reactions Facebook allows – love, wow, haha, sad, angry and care – Trump has not received more angry reactions than he did in 2016. Instead, he’s increased the proportion of people who “loved” his posts between 2016 and 2020.

Although we have seen an increase in comments (which may be negative), when examining reactions (Love, Wow, Haha, Sad, Angry and Care) in Graph 5, it is clear that Trump has not received more Angry reactions, but instead more Love. This is in contrast to Biden, who has received less Love and Angry reactions, but more Sad reactions when compared to Hillary Clinton.

Graph 5. Types of reactions for the pages (Care did not exist in 2016)

Content form – The rise of live video and decline of photo

Content approaches are ever evolving, with the use of video a recent phenomenon that has taken over almost all other content forms. Today, parties and candidates have (through Facebook) in effect their own Cable TV stations. This potential is reflected in the pages’ ever increasing use of Facebook Live video and other video approaches as seen in Graph 6. This rise in video has seen the use of links (such as to news websites) or photo content (such as posters) rapidly decline between the elections. Trump’s page has been the most interested in adapting and changing approach, while Biden has kept hold of more traditional forms such as photo. This again shows the Trump campaign evolving their approach in a more concerted way.

Graph 6. Content type of the sent posts

Total post views – Are the candidates messages spreading?

Finally, in order to appreciate total reach, we can use average video views as a proxy because they auto-play after 1 second. In Graph 7 we can see that Hillary Clinton performed very well for the engagement she received, in 2016 she gathered 423,429 views per video post to Trump’s 346,532. Her audience was extremely effective in their sharing, and content was clearly engaging. However, today Joe Biden has fallen behind badly with 130,737 views per post to Trump’s 441,722. Although Biden has an engaged audience, his messages appear to not be getting out into the Facebook population. Biden’s small active user-base is failing to deliver content into people’s news feeds and thus rack up views. This spells out serious problems for the Biden campaign, and the necessity of relying on targeted advertising. Despite how discussed targeted advertising is, I believe the real power of Facebook lies in the mixing of political messaging with real people acting as the messengers. People trust their friends and families not politicians, thus making shared organic political content more impactful.

Graph 7. Average video post views and overall total views

Conclusions

The data is presenting a remarkably similar picture to 2016. Today, Facebook is still the most important social network because it is more representative of the US population and people’s Facebook networks are more real-world based than network’s such as Twitter. As such, it is my opinion that the impact of the organic campaign will likely outweigh the impact of targeted advertising. Therefore engagement matters, and it is in this popularity contest that we see Trump succeeding once again.

Overall, the data fits into what many claim is Facebook’s overtly right-wing skew. However, as the 2016 data shows, Clinton was performing relatively well for her audience size. It is instead apparent that the Biden campaign is failing to make an impact on the platform. This is not a given as both Clinton and Bernie Sanders created successful campaigns that reached out to the wider public through supporters. If the Biden campaign fails to battle Trump upon Facebook, instead focusing on more liberal, younger networks such as Twitter or Snapchat, or relying solely on Facebook adverts, the Democrats are making a misstep. Biden’s new approach to stop all negative advertising on the platform will also likely not help the situation as users are drawn to content that incites reaction. While Facebook’s advertising one-week pre-election hiatus will only help Trump.

It is possible for Trump to run a successful Facebook campaigning and still lose, see for example Jeremy Corbyn’s excellent 2017 General Election campaign. However, Facebook engagement is important, because through activities such as sharing, individuals are rebroadcasting party content in a more effective way than previous tools allowed for. Today a wide mass of people, likely non party members, are acting like party digital doorstep campaigners, delivering political content integrated with their own social characteristics. And at the moment on Facebook more people are doing this for Trump than for Biden.


Data via Crowdtangle, all analysis undertaken by Tristan Hotham .


0 comments on “Biden’s organic Facebook campaign is behind Clinton’s, while Trump is surpassing his 2016 performance

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: