The Facts Behind Allegations of Political Bias on Social Media
Social media has transformed politics, allowing ordinary Americans to connect more easily with elected officials and candidates, access up-to-the-moment news, and share their political opinions with millions of others. It has also allowed candidates and campaigns to connect directly with voters without the filter of traditional media. These changes have thrust social media platforms into the center of political debate, with candidates or campaigns often blaming social media for political bias when they lose or arguing that social media is “destroying democracy.” As the United States enters another heated presidential campaign cycle, policymakers need a firm grip on the facts about political bias on social media to avoid changes that could do more harm than good.
Both sides of the aisle, but most frequently Republicans, accuse social media companies of displaying bias in their content moderation and news feeds. There is some research into these alleged phenomena, but not enough, and much of the current evidence to support this view relies on anecdotes and scant research. Before policymakers jump straight to regulating social media to address alleged bias—and likely running into First Amendment issues—they need answers to several key questions, namely: Is there political bias on social media? What does this bias look like? And how does it affect American politics?
What Is the Evidence for Anti-conservative Bias on Social Media?
Those who believe there is anti-conservative bias on social media platforms cite plenty of anecdotal evidence to back up their claims. Multiple right-wing political activists and commentators have faced social media suspensions or permanent bans for offenses ranging from election or COVID-19 misinformation to hateful conduct—or even in error. Notable cases include MyPillow CEO Mike Lindell, former national security advisor Michael Flynn, former White House chief strategist Steve Bannon, and media personality Candace Owens. Of the American elected officials who have been suspended or banned from social media, all are Republicans.
Most famously, former President Donald Trump was banned from Facebook, Twitter (now X), and YouTube, among other platforms, in January 2021 for violating the platforms’ policies on inciting violence. Many Republican politicians and conservative media outlets criticized the bans. Even outside strictly conservative spaces, the bans sparked debate over how much power social media companies should have to censor democratically elected political leaders. Twitter eventually lifted its ban in November 2022, Facebook followed suit in January 2023, and finally, YouTube lifted its ban in March 2023.
In addition to suspensions and bans of prominent conservatives, critics argue that mainstream social media platforms censor conservative speech to promote a liberal political agenda. Proponents of this view often point to Twitter blocking users from linking to an October 2020 New York Post article about the contents of a laptop that once belonged to Hunter Biden, son of then-presidential candidate Joe Biden. Then-CEO of Twitter, Jack Dorsey, admitted that this decision was “wrong,” and Twitter implemented changes such as labeling potentially misleading tweets instead of removing them.
Allegations of anti-conservative bias on social media have led policymakers to introduce multiple bills that would amend Section 230 of the Communications Decency Act, a much-debated law that protects online services from liability for removing or failing to remove content. These bills would limit the types of content social media platforms could remove without facing potential lawsuits. At the state level, both Florida and Texas have passed laws aimed to curtail alleged anti-conservative censorship on social media. Technology industry groups challenged these laws on First Amendment grounds, and the cases have made their way to the Supreme Court.
Judging by their political donations, the vast majority of employees of social media companies such as Facebook (now Meta) and Twitter support liberal candidates, which many Republicans argue is the reason these companies are biased against them. However, donations from companies’ political action committees (PAC) are more balanced, donating predominately to incumbent candidates regardless of the candidates’ party affiliation.
There is limited research into whether mainstream social media platforms display anti-conservative bias, though what research exists does not support these claims. One study by NYU Stern did not find evidence of such bias, concluding that claims of bias were “based on distortions and falsehoods.” On the contrary, it found that conservative politicians and media outlets were often more successful than their liberal counterparts at generating social media interactions. However, critics of the study pointed out that it was funded by Craig Newmark, who has donated to Biden’s campaign, the Democratic National Committee, and other liberal groups and causes, calling the impartiality of the study into question.
In 2016, prompted by concerns of Republican Senator John Thune—then-Chairman of the Senate Committee on Commerce, Science, and Transportation—Facebook launched an investigation into allegations of political bias in its Trending Topics feature. The investigation found “no evidence of systematic political bias in the selection or prominence of stories,” though an internal investigation conducted by a company accused of bias is unlikely to convince detractors. Additionally, a 2021 study published in Nature journal by researchers from Indiana University found “no strong or consistent evidence of political bias” in Twitter’s news feed. However, this does not settle the issue of whether Twitter’s content moderation is biased.
What Is the Evidence for Pro-conservative Bias on Social Media?
Some critics argue that social media has a pro-conservative bias. Most often they argue this bias is displayed as insufficiently policing hate speech, misinformation and disinformation, and other forms of content that many on the left find objectionable. However, some individuals and organizations argue that rather than being biased against conservatives, mainstream social media platforms actually amplify conservative users and content above their liberal counterparts.
The progressive non-profit research center Media Matters has published multiple studies of engagement on Facebook pages that regularly post about American political news, finding in 2019 that right-leaning and left-leaning pages earn similar interactions and, in 2020, that right-leaning pages earned more interactions. To be specific, between January 1 and September 30, 2020, right-leaning Facebook pages earned 43 percent of total interactions despite accounting for only 26 percent of posts, although this may in part be due to left-leaning pages being less active and posting less.
There is some additional evidence to support claims that social media favors conservatives, either unintentionally or by design. A 2022 study by researchers from the United States and Italy looked at social media activity, web browsing, and news sources surrounding the Black Lives Matter protests in June 2020. It found that right-leaning content was viewed and shared “substantially more” than left-leaning content on Twitter. According to the study, right-leaning news outlets achieve higher visibility and engagement on social media, giving them an advantage. This could be a result of biased algorithms, but it just as easily could be a result of user behavior.
In 2021, Twitter itself examined tweets from elected officials in the United States, United Kingdom, Canada, France, Germany, Spain, and Japan and found that in all of these countries but Germany, tweets from right-wing politicians were amplified more than those from left-wing politicians, and this effect was greater on an algorithmic timeline than a chronological one. Likewise, right-leaning news outlets were amplified more than left-leaning outlets.
Facebook claims that conservatives’ advantage on social media comes down to the tactics right-wing populist figures use being best suited to the platform. Many of these figures create content that draws on users’ emotions, particularly anger and fear, which drives engagement. Some left-leaning sources disagree with this claim, arguing instead that Facebook’s algorithm promotes conservative sources by design. Progressive magazine Mother Jones cites a Wall Street Journal article alleging that, when Facebook made changes to its newsfeed algorithm in 2017 to minimize the presence of political news, engineers deliberately tweaked the changes to disproportionately impact left-leaning news outlets.
Why Does This Matter?
The left and right agree that there are problems with how mainstream social media platforms moderate content, but they disagree on what those problems are. The question of whether these platforms are biased either in favor of or against conservative content is a key point of contention. Social media is an important part of modern politics, and if any mainstream platforms are biased—in their content moderation decisions, news feed algorithms, terms of service, or any other number of ways—it would have far-reaching implications. As the 2020 U.S. presidential election demonstrated, even the suggestion of bias in any part of the democratic process can cause significant political upheaval.
In addition to the potential impact on elections, bias on mainstream social media platforms would have a direct impact on individuals and organizations that rely on those platforms as a source of revenue. This includes news outlets, who benefit from advertising and subscription revenue gained from social media users discovering their content. It also includes content creators on platforms such as YouTube, where creators earn a percentage of the advertising revenue from their videos.
Both right- and left-leaning YouTube channels have complained of having their political content demonetized. On the one hand, YouTube relies on advertising and understandably exercises caution when it comes to placing ads next to content that advertisers might find objectionable, including political content. On the other hand, this dynamic makes it difficult for creators to make content addressing controversial political topics without potentially threatening their livelihood. In addition, creators complain that the YouTube monetization system lacks transparency, making it difficult to predict whether a video will be demonetized.
There is some research into whether mainstream social media platforms are politically biased, but not enough to reach a conclusion that proves bias one way or another or disprove allegations of bias entirely. Contributing to this problem is a lack of insight into how content moderation takes place on these platforms. Platforms such as Facebook have begun to address the problem by making some datasets available to researchers, which hopefully represents a new trend toward greater transparency that will enable further research into key issues like political bias.
In the meantime, policymakers on both sides of the aisle should avoid jumping to conclusions based on anecdotal evidence, which leads to laws like those in Florida and Texas that create more problems than they solve. Policymakers should also look for areas where the left and the right can agree on the problem and are therefore more likely to reach a compromise on the solution, such as child sexual abuse material and state-sponsored disinformation. Additionally, policymakers should push for more research into the question of political bias on social media and greater transparency in platforms’ content moderation outcomes.
