In the past couple of weeks, we’ve seen tech companies and civic organizations step up their efforts to provide reliable sources of civic information on their platforms. In response, social media has blown up with voting content, much of it in the form of familiar pictures, such as family members and friends proudly showing off their “I voted” stickers on Facebook and Instagram. Our digital spaces look more and more like they’re meant to support the democratic purpose, but are they really doing enough? Can they accommodate the principles of a democracy before it’s too late?

The efforts of malicious actors (and the complicit individuals who enable them) in our digital spheres have been well documented; their success in manipulating flows of information has been unprecedented. Although many platforms are pushing to get out the vote, misinformation and confusion surrounding the election persist online. As Election Day approaches, tech companies and politicians must each minimize their contributions to these existential threats.  

The most obvious step that we can take is to address the recent flood of misinformation, conspiracy theories, and calls for violence on social media, particularly as they emerge from individuals and politicians with influential platforms. Tech companies have the responsibility to continue flagging this content with warning labels and releasing explicit statements that address what types of content are allowed on their platforms. For too long, companies like Twitter and Facebook have wavered in their commitment to moderate problematic sources of digital content. In a time as crucial as the leadup to the 2020 elections, we need to know that these companies will stand firm against the viral spread of harmful content on the Internet, rather than enabling it. 

The responsibility to make the Internet an informative space lies not only in the tools that we use, but also in our individual actions. Spreading blatant lies and posting harmful content have become accepted norms of social media discourse. The harm caused by this behavior is amplified by people like Donald Trump who have digital followings in the tens of millions. Too many politicians are denigrating our democracy by promoting conspiracy theories and misinformation and, in extreme examples, even calling for violence. Politicians today simply do not understand that promoting conspiracy theories and misinformation and calling for violence denigrates an informed democracy. It’s time that we demand more from our politicians and how they behave online, as their digital platforms reflect not only their individual stature, but also our democracy as a whole. In the days leading up to the election, politicians must be expected to adhere to better standards of conduct on the Internet. Refusing to spread any false information pertaining to the elections and the COVID-19 pandemic is a baseline. A harmonious union of technology and democracy can only occur if political figures are committed to fostering civil discourse and sharing reliable information on their platforms.

More voters this year will be mailing in their ballots than ever before, but states do not have the resources necessary to count them efficiently. In most states, including Michigan, California, and Texas, election workers cannot begin counting ballots until Election Day. In several states, including Pennsylvania and Wisconsin, election workers are not even allowed to begin processing ballots (opening envelopes, verifying signatures, etc) until Election Day. These rules, along with dated technology and the influx of mail-in ballots, mean that determining the result of the election will take much longer than usual.

Some social media platforms already have clear policies in place. Facebook and Twitter plan to apply informational labels to posts that declare a winner before the election is authoritatively called. These efforts, though hopefully effective, will not be enough to acclimate Americans to the possible delay in results this November. Journalists and news networks, too, ought to exercise caution in their reporting of the results. Early leads may mean less this year, and should be reported as such. When reporters share what percentage of precincts are reporting, they must be careful to let viewers know whether that percentage takes mail-in ballots into account. All the sources upon which Americans rely for accurate election results have a great deal of responsibility this year. They must exercise caution and patience in calling all the elections on the ballot.

But the rest of us have responsibilities, too. We need to be prepared for a long wait as election workers across the country diligently count our ballots. Their work is integral to our democracy. It would be a disservice to them to dispute election results simply because of a delay. Candidates with an early lead may lose. Candidates with a strong showing among the votes cast in person may lose. A race may seem close, and then become a landslide. It may seem like a landslide, and then become close. None of these possibilities suggests that the results are fraudulent or mistaken. We need to be wary of sharing and celebrating unverified results. And when verified results are available, we need to be ready to accept them.

Regardless of who wins the election, these past four years have demonstrated the breadth of issues raised by technology’s integration into our private and public lives. We’ve learned that social media creates a platform for both benign and malicious actors to gain influence and exert power. On the Internet, users can distort and rewrite reality in the form of rumors and falsehoods, which blur the line between fact and fiction. This simulated version of reality has another flaw — it is hackable. The constant evolution and improvement of technology enables ill-intentioned actors to enact violence and harm reality through the co-optation of online networks, which we have already seen in the form of cyber-attacks, extremist propaganda, deep fakes, and more. The Internet also allows bad actors to evade responsibility by using the veneer of a screen to cover their hate and lies. 

Political figures at the highest level exploit this system for personal gain, sowing doubt and anger in their followers. While these individuals are certainly at fault for cultivating toxic online environments, they are simply taking advantage of systems that reward sensational and polarizing content with followers and clicks. The pre-existing conditions of social media actually incentivize the spread of misinformation, as long as the content elicits strong emotions and aligns with the worldviews of followers. Misinformation undermines reality, fuels misunderstanding, and sows doubt in systems, like the voting process, which we rely on to make democracy work. 

Social media inherently creates a favorable environment for the formation of polarized groups, which often facilitate the spread of misinformation. Political scientist Robert Putnam reports in his book, Bowling Alone, that while “real-world interactions often force us to deal with diversity, the virtual world may be more homogeneous.” This occurs because of users’ confirmation bias, their tendency to prefer information that reinforces their own beliefs. In practice, approximately 75% of retweets on political topics take place between users of similar ideological views. This phenomenon is exacerbated by recommendation algorithms that promote content that users already agree with and limit their exposure to content that challenges their worldview.

A less obvious consequence of the rise of social media is the changing nature of the “mediators” of information. Before social media was popularized, people like journalists and reporters created and disseminated information, serving as touch points between the news source and the consumer. Social media’s user-to-user networks removed the need for an intermediary actor, so news and information can now be manufactured and shared by consumers themselves. Today, articles from the Wall Street Journal appear side-by-side on Facebook with the rantings of hate or conspiracy groups. The potential to selectively view and share information that strongly reinforces users’ viewpoints increases the likelihood of confirmation bias, echo chambers, and the formation of ideologically homogeneous groups, all of which contribute to societal division. 

The Internet is not the root cause of polarization in our country today. There is a web of interconnected historical and societal factors that contribute to the national atmosphere of division which cannot be resolved solely through the reformation of the Internet and social media. However, it would be irresponsible to deny or downplay the crucial role that they play in normalizing and amplifying the polarization that has led to such hate and misunderstanding in our country.

Rethinking the Platform

Social media platforms function as conduits for polarization and misinformation, bringing harm to our democracy and our society. In order to move towards an Internet free of these threats, tech companies must consider the ramifications of their platforms’ structures and enact significant change.

Rethinking recommendation algorithms is a good starting point for tech companies when it comes to reducing polarization. Renee DiResta, a technical research manager at Stanford Internet Observatory, suggests that companies develop a transparent list of “Do Not Amplify” topics that will not be recommended to their platforms’ users, which could include conspiracy theories and content promoting violence. Recommendation algorithms can also be used to actively combat polarization, as demonstrated by a Google Jigsaw project that showed anti-ISIS content to potential ISIS sympathizers alongside ISIS-related search results. A similar strategy could be used to offset other extremist content, and even to add balance to content that is ideologically polarizing but not extremist.
To minimize the harmful effects of misinformation, companies should build more robust systems for reporting and flagging misleading content. For example, there is evidence that crowdsourced fact-checking works, and it could expand platforms’ fact-checking capabilities beyond the limited reach of their employees. Twitter is already developing a new feature, Birdwatch, which will allow users to submit detailed reports about potentially misleading content. In addition, platforms should more consistently and clearly flag misleading posts (for example, posts that have accumulated a certain number of reports about misinformation, or a certain proportion of reports relative to the number of views), as well as accounts that have been frequently reported for misinformation. Misleading posts and frequently misleading accounts should not be amplified by recommendation algorithms.

Taking Actions as Individuals

Ultimately, the onus should be on tech companies to fundamentally change their platforms so that they no longer facilitate polarization and misinformation. In the meantime, however, individuals can promote healthier experiences for themselves within these platforms’ existing, flawed frameworks. This starts with personal responsibility. We may not control the underlying algorithms and structures of the platforms we use, but we can control the information we put out there. We must each avoid becoming part of the user-to-user spread of misinformation by staying informed through credible news sources, unfollowing frequently misleading accounts, and critically reading any articles or other content we want to share. We should also practice kindness and courtesy in all of our online interactions, rather than contributing to the toxicity that gives rise to polarization.

If we recognize and understand how recommendation algorithms can influence us, we can take steps to fight our own confirmation bias and avoid polarization. For example, following credible journalists and news sources from across the political spectrum introduces more balanced content into our feeds. Similarly, “liking” a broad range of content prevents recommendation algorithms from developing a narrow, biased profile of our opinions. If we sort our feeds by “most recent” whenever possible, such as on Facebook and Twitter, we won’t encounter a disproportionate quantity of recommended content. We should also avoid unfriending users over differences of opinion, since this makes our virtual worlds more polarized. Finally, we should be mindful of influencers (users who have more followers than they are following), since their content is amplified across social networks and can contribute excessively to polarization. If an influencer only engages with other content and users that reinforce their opinions, we should consider unfollowing them.

There are many steps that we can take to change how we consume and spread information on social media platforms. Perhaps the most radical one is stepping away. Social media is an imperfect representation of the real world, distorted towards sensationalism, misinformation, toxicity, and polarization. If we choose to exist in our virtual world less and our real world more, we can avoid experiencing and contributing to all the negativity online.