In this final blog post from a series of four, CDE member Dr Gillian Bolsover, focuses her analysis to the impact of COVID on the US 2020 election, presenting initial insights from her analysis of social media discourse in the early campaign period.
Responding to the global challenge of the COVID-19 pandemic, myself and my team of research assistants have been busy over the summer producing research about how the disease is impacting many facets of life. To mark the beginning of term, we are now publishing our fourth and final data memo in the COVID series. This data memo analyses discourse on US Twitter to understand how the coronavirus pandemic is affecting the upcoming Presidential Election that will take place in November.
In a typical election year, media discourse over the summer months would be dominated by campaign coverage. However, across the world attention has been focused on COVID-19. As of the end of August, more than 88 elections across the world have been delayed due to concerns about disease transmission. In the US, numerous presidential primaries were delayed (although the two major parties’ nominees were largely decided before these cancellations). The extent to which candidates are changing their campaigning in light of COVID already demonstrates that the disease has become a political issue. For instance, Democratic candidate Joe Biden has announced he will no longer hold rallies while COVID continues to spread in the US. In contrast, Republican incumbent Donald Trump continues to hold campaign rallies, some of which have been implicated in disease outbreaks in rally areas.
Whether greater postal voting is or could be introduced in the US to facilitate social distancing is another point of contention. Regulations vary widely between US states with five conducting ballots almost entirely by mail and many moving toward allowing all voters to request a mail-in ballot in the context of COVID. However, some other states are resisting these reforms, most notably Texas. Texan Attorney General announced in April that fear of contracting COVID would not be considered a valid reason for requesting a postal ballot in the state. In July, the US Supreme court refused to expedite a case by Texan Democrats contesting the Texas Attorney General’s decision, meaning that it is unlikely that the country’s highest court will hear the case before the November election.
This resistance is likely linked to Republican Party claims (unsupported by research) that mail in voting has high rates of fraud and allegations that Democrats will use fraudulent mail in ballots to “steal the election.” This argument was advanced by President Trump in an August tweet that read:
So now the Democrats are using Mail Drop Boxes, which are a voter security disaster. Among other things, they make it possible for a person to vote multiple times. Also, who controls them, are they placed in Republican or Democrat areas? They are not Covid sanitized. A big fraud!
With social media now constituting a major venue for political campaigning, news reading, information dissemination and political discussion, our recently published data memo analyses a random selection of tweets in trends about the election during two weeklong periods: 25 – 31 May and 6 – 12 July. These periods both represent different stages in the fast-moving, socio-political climate in the US of what is normally the early campaign season.
In each weeklong period, we randomly selected 250 tweets for analysis from trends about the election. Of these, only a small number considered both COVID and the election. During the May period, a number of these tweets repeated the allegations of postal voter fraud and Democratic beneficiaries of postal voting advanced by Trump, as well as mirroring Trump’s discursive style. Several tweets advanced other conspiracy-style content, such as that Democratic contender Bernie Sanders was threatened into dropping out of the race. Other posts referenced sources to make claims that were unsupported by the original source.
In the July period, we saw a heavy influence of the viral videos produced by a particular Political Action Committee (PAC) that was formed to oppose the re-election of Donald Trump. Of all 299,347 posts in 31 hashtags related to the election in the 6 – 12 July period, 2.5% contained one of the two hashtags initiated by the PAC’s viral videos released during the week. This demonstrates the enormous influence that these campaign finance pooling organisations have over elections in the US.
Along with these viral videos, we saw more criticism of Trump in the July random sample of posts about COVID and the election compared to the May sample. However, this difference was not mirrored in wider election discourse. This is likely because Biden, the Democratic Party and left-wing voices have approached the pandemic much more seriously in the US, with Trump, the Republican party and right-wing voices consistently arguing that the severity of the disease is being overstated and that a change in behaviour is neither necessary not warranted. As such, it is reasonable to hypothesise that it would be left-wing voices that were more active in discussing COVID in relation to the election.
Within the wider sample of tweets about the election in election trends and the broader category of tweets about politics in election trends, the majority of the focus was on Trump, with twice as much discussion of Trump as of Biden. This could well be positive for the Trump campaign as research has demonstrated that candidate visibility is a core component of election success separate to candidate favourability. Although for both candidates the number of critical comments was higher than the number of positive comments, the ratio of critical comments to positive comments was higher for Biden than for Trump in each sample.
What was clear in the analysis was that the different parties’ approaches to COVID have become a major campaigning position, in contravention of the non-political, public health-orientated approach that that would be desirable for pandemic policy and information dissemination. This analysis suggests that partisan voices, including the President, have been successful in making COVID a polarised political issue. Thus, the question we must ask going forward is, unfortunately, not how will COVID affect the US 2020 election but how will COVID be used as a political issue in the US 2020 election.
The re-casting of the COVID pandemic as a political identity issue means that polarised positions become fixed and not amenable to debate. This opens the door for a greater amount of misinformation, hyperpartisan content, conspiracy theory and PAC influence that aim to leverage and enflame entrenched positions on this fast-evolving public health issue for political advantage.
A more detailed analysis can be found in the data memo.
This is the final post from a series of four by Dr Bolsover – you can find the previous three posts at: