Narratives Targeting Electronic Voting Machines

Authors: Jack Nassetta, Graphika and Jack Cable, Stanford Internet Observatory

Contributors: Sahar Markovich, Emily Ross, Shelby Perkins, Chase Small, Nazli Ugur Koyluoglu, Dylan Junkin, Isabella Garcia-Camargo, Stanford Internet Observatory; Andrew Beers, Kolina Koltai, Ian Kennedy, University of Washington Center for an Informed Public; Ben Nimmo, Graphika


The narrative that electronic voting machines could be compromised in an effort to change the outcome of the election has been a recurring theme on social media over multiple election cycles. In recent weeks we have seen the narrative crop up among the left and right that certain voting machines could not be trusted for various reasons. These claims seek to delegitimize the voting process for the millions of voters who will be using electronic voting machines on Election Day. This narrative often exploits misconceptions around electronic voting, a general fear of foreign or domestic hacking and unfalisable rumors. So far we have identified three key sub-narratives to watch.

A Deep-Rooted Fear

Following the 2016 election cycle, 55% of Americans were “not too” or “not at all” confident that America’s election systems were secure from hacking. Additionally in 2020, 63% of Democrats, and 26% of Republicans were extremely or very concerned that foreign countries would tamper with election systems. It is therefore not surprising that claims of hacking or tampering with voting systems are so often seen on social media. Untrue claims about voting machines manufacturers feed underlying fears about election security and can quickly go viral.

Although voting machines can certainly have vulnerabilities, attacks are mitigated by various security controls. This year, election experts estimate that 95% of votes cast will have a verifiable paper trail. When combined with post-election audits — such as risk-limiting audits — election officials and the public can verify that tabulated results match the paper votes. Thus, any large-scale tampering of voting machines with a paper trail will be detected.

Narrative 1: Voting Machine Ownership

A repeating theme in the narratives delegitimizing voting machines suggests nefarious ownership of a company that produces voting machines and that machines are being rigged as part of the manufacturing process. On the right in 2016 and 2018 this narrative used George Soros as the supposed owner of voting machine company Smartmatic. This claim spread on Facebook and Twitter while being repeatedly debunked. In 2020, we have seen a parallel narrative emerge on the left concerning Ivanka Trump being issued a patent for voting machines by China. Despite no evidence that these voting machines would ever be used in the United States, or anywhere, a narrative spread on Twitter and Facebook that voters should be concerned. 

Screen Shot 2020-10-20 at 12.34.09 AM.png
Screen Shot 2020-10-20 at 12.34.51 AM.png

Narrative 2: Tampering with Machine Programing

An additional narrative we observed around the legitimacy of voting machines is concern that their programming for vote tabulation could be tampered with. This concern reached a peak in recent weeks after a story emerged that memory sticks used to program Philadelphia voting machines were stolen. Despite city officials confirming that the incident would not compromise the integrity of the election, Twitter users quickly weaponized the story in order to cast doubt on the election. 

Screen Shot 2020-10-20 at 12.35.34 AM.png
Twitter users responding to reports that the memory sticks were stolen.

Twitter users responding to reports that the memory sticks were stolen.

Narrative 3: Voting Machines Changing Selection

The final major narrative we have seen so far this cycle is again one that has been seen previously, that voting machines are changing voter selections. This concern is based on legitimate instances of this happening, but the narrative has extended to unsourced rumors, including amplification in 2016 by Russian actors. On Election Day 2016 the Russian Internet Research Agency-controlled “Tennessee GOP” Twitter account spread the narrative that voters were unable to select the Trump/Pence button on an electronic voting machine in an attempt to claim that it was rigged for Hillary Clinton (Link to archive of TEN_GOP activity via NBC News). In the last week we have seen a similar unverified narrative from an anonymous source appear and be spread across Twitter and Facebook. In this case, it was the claim that an electronic voting machine was switching any selection of Biden/Harris to Trump/Pence and that an election judge subsequently ordered the machine shutdown. This unverified story was removed by Twitter but continued to spread as it was copied word for word by other accounts, casting doubt on election integrity. 

Screen Shot 2020-10-19 at 8.32.41 AM.png
unnamed (7).png

Takeaways

Concerns and fear regarding the security of electronic voting machines are not new to this election cycle and are unlikely to go away. The danger comes when these fears are exploited by claims that are often unsourced. Claims such as “a friend told me their vote was changed” cannot be fact-checked and yet spread virally delegitimizing the voting process. In the 2020 election cycle, we have seen the major platforms handling this differently, with Twitter taking down the above posts about vote switching and Facebook not. We expect increased activity around this narrative as Election Day draws closer and more people cast their vote in person. Platforms should establish consistent policy to remove unverifiable content that casts doubt on the integrity of the voting process.

Previous
Previous

Platforms of Babel: Inconsistent misinformation support in non-English languages

Next
Next

EIP Weekly Update: Oct. 20