Facebook and Democratic Elections Simply Don't MixJordan Guiao / Mar 25, 2022
With a federal election anticipated in Australia within weeks, Meta– which operates Facebook, Instagram and WhatsApp– has announced initiatives it claims will help protect the integrity of the elections and stem the tide of disinformation while bolstering the security of Australian politicians during this period. The announcements include more investment in fact-checking partners, a media literacy campaign for the public, and a review of security protocols targeting politicians in the platform – all taken by themselves, welcome additions. But these piecemeal efforts obscure the central issue with Facebook – that its platform and business model is incompatible with democratic and civic processes like elections.
Controversy follows Facebook from the last Australian federal election in 2019. In that cycle, a widely shared false rumor about the opposition Labor Party generated vast traction, doing significant reputational damage and likely costing votes. Foreign influence operations and coordinated inauthentic behavior muddied the public discourse, while Facebook ads were microtargeted to Australians who are fans of former President Trump.
Of course, it’s not just in Australia where such things take place. Facebook’s track record with elections is full of tragedy. Even Noble Laureate Maria Ressa’s incredulous reaction to the Australian announcement is indicative of how damaged Facebook’s reputation is when it comes to its role in democratic elections.
For instance, consider the now infamous 2018 Cambridge Analytica incident, in which the data company used stolen Facebook data to target voters for the 2016 Trump campaign, applying psychographic targeting and manipulation at an industrial scale. Despite the high-profile, headline grabbing nature of the scandal and some large (and not so large) fines from regulators around the world, Facebook’s promised audit of its platform for any other potentially harmful apps has still not surfaced, there has still been no proper investigation into the practices which facilitated the Cambridge Analytica issue, and still no legislation or ruling that strikes at the heart of these privacy abuses.
Whistleblower and former Facebook integrity team data scientist Sophie Zhang revealed in late 2020 how Facebook has continually been used to undermine elections around the world. The litany of offenses across different countries is long:
- In Honduras, inauthentic assets were usedto artificially boost former President Hernandez and deceive local citizens. It took Facebook months to act on this, by which time the damage was done, and even after action had been taken, the activity returned shortly afterwards.
- In Brazil, there were over 10.5 million fake reactions and followers removed from high profile politicians during past election periods.
- In Azerbaijan, the ruling political party used thousands of inauthentic assets to undermine the opposition, which Facebook barely acted on.
- In Bolivia and Ecuador, Zhang discovered instances of inauthentic activity and coordinated engagement from bots, but could not act on these due to her workload; and so, it was simply ignored.
- In India, there was evidence of a sophisticated propaganda network of more than a thousand actors aiming to influence local elections in 2020.
A year later, another whistleblower named Francis Haugen rocked the world by leaking tens of thousands of internal Facebook documents demonstrating just how troubled Facebook’s approach to elections and civic integrity truly is.
Haugen worked as a product manager in the civic integrity team, which looked at election interference globally, among other topics. She grew increasingly frustrated at Facebook’s lack of action on the issues her team came across. She revealed that the civic integrity team had many proposals to address harmful problems on Facebook, but those initiatives were often blocked by Mark Zuckerberg directly.
This dynamic played out disastrously in the United States in the 2020 presidential election and in its aftermath. In the runup to election day, Facebook announced its most comprehensive effort to date to protect the election. The company announced sweeping activities aimed at improving the information environment, and imposed stricter rules for advertising and political campaigns. There were increased efforts to remove voter suppression related content, and extra content labelling for potentially misleading posts. Facebook even prohibited political advertising in the final days leading up to the election.
But these steps were ultimately insufficient. A report by the global activist group Avaaz analyzed Facebook’s interventions during the US 2020 elections and found that the company did not act quickly enough to suppress false and toxic content, a fact that created “the conditions that swept America down the dark path from election to insurrection” leading to the Capitol Hill riots, with the top 100 misinformation spreaders receiving millions more interactions on Facebook than credible news brands. Frances Haugen also revealed that Facebook was well aware of the behavior on the platform that contributed to the January 6, 2021 riots and Capitol Hill, and that the company did not have sufficient plans or resources in place to thwart the Stop the Steal movement using its platform to spread election lies.
Of course, these various failures were only reported because of Haugen and Zhang’s decisions to come forward. Global case studies are less well documented than US or European ones, but we must not forget these examples, for which Facebook has still made no real reparations.
If Facebook cannot adequately address election misinformation on its home soil, what hope does the rest of the world have – with its tiny satellite offices mostly set up to sell more advertising? Certainly, a few extra dollars to fund fact checkers in Australia isn’t going to make a material difference.
When will we finally accept that Facebook and elections– indeed, Facebook and democracy– simply do not mix?