No, it’s not proof of “bias” when social media companies enforce their rules
Conservative media learn the hard way that rules disproportionately affect rulebreakers
On Monday, the right-wing Media Research Center released a study aimed at proving the long-held yet long-debunked idea that social media companies have an anti-conservative bias.
“Big Tech has caused serious damage to President Donald Trump’s ability to be heard on social media,” claimed MRC. The evidence, according to them, is that Facebook and Twitter have never censored the social media accounts of former vice president and Democratic presidential nominee Joe Biden. Trump, on the other hand, has had his accounts censored 65 times over the past two years. The Federalist, Fox News, The Washington Times, and the Christian Broadcasting Network were among the pro-Trump media outlets that lept at the opportunity to declare vindication after years of banging the “anti-conservative bias” drum despite a complete lack of evidence.
Unfortunately for MRC and the right-wing outlets celebrating this study, it’s not actually evidence of anything other than what has been pretty clear all along: Trump frequently breaks Twitter’s terms of service. But the takeaway from MRC’s study is that anything less than an even split between the two candidates in terms of Twitter enforcement is evidence of bias. That would be true if Biden and Trump both used Twitter the same way, but we all know they don’t.
The idea that this is evidence of bias is laughable on its face. On Twitter, author Dennis DiClaudio highlighted the absurdity of this “gotcha” study, tweeting, “My dog has been punished for peeing on the carpet 65 times, my wife zero.”
Each example of what MRC calls “censorship” is actually just the company enforcing its existing policies.
The examples of “censorship” provided by MRC in its blog post on the study are laughable. When Trump falsely claimed that “there is NO WAY (ZERO!) that Mail-In Ballots will be anything less than substantially fraudulent,” he was violating the site’s civic integrity policy. It’s a lie to say that mail-in ballots are “fraudulent,” and the president knows this, being someone who regularly casts his vote by mail. While Twitter did not censor these tweets, it did append a note to the bottom linking to information about mail-in voting.
When Trump twice shared a badly edited video showing Biden played N.W.A.’s 1988 song “Fuck Tha Police” on his cell phone at an event, Twitter added a “manipulated media” tag to the clip. In reality, Biden was playing a snippet of the song “Despacito.” This makes sense in context, as the song’s creator, singer Luis Fonsi, had introduced Biden moments earlier.
When Twitter removed a video Trump retweeted featuring Dr. Stella Immanuel, who believes that the government is run in part by “reptilians” and that alien DNA is being secretly used in some modern medicine, conservatives cried foul. Immanuel’s video called the drug hydroxychloroquine a “cure.” This was a clear violation of Twitter’s rules on spreading false information about COVID-19, which restricts statements of fact about “preventative measures, treatments/cures, and other precautions.” Had Immanuel’s claims been phrased in the form of an opinion or a hypothesis, it wouldn’t have violated this policy.
Yet another example listed by MRC as Twitter censoring content was when the company removed a video featuring the song “Everybody Hurts” by R.E.M. because Universal Music Publishing Group, which controls the copyright to the song, had issued a takedown request for unauthorized use. Per the Digital Millennium Copyright Act, platforms like Twitter and YouTube can’t be held liable for copyright infringement so long as “the service provider promptly removes or blocks access to the material identified in the notification.” In short, when a copyright holder files a takedown request with Twitter, the company has a legal obligation to remove it. This is not partisan bias; it is the law.
One of the MRC-listed examples of “censorship” on Facebook centers on Facebook’s decision to deactivate a Trump campaign ad featuring Nazi iconography historically used to label political prisoners in concentration camps. Facebook removed the campaign’s 88 versions of this ad. Perhaps coincidentally, according to the Anti-Defamation League, “88” is an extremely common white supremacist symbol meant to signal “Heil Hitler,” as “H” is the eighth letter in the alphabet. “Our policy prohibits using a banned hate group’s symbol to identify political prisoners without the context that condemns or discusses the symbol,” said a Facebook company spokesperson at the time.
If anything, Trump’s continued use of social media is evidence of a pro-Trump bias from Facebook and Twitter.
Throughout Trump’s first two and a half years in office, the president seemed unwilling to adhere to the same policies the rest of Twitter users were expected to follow. In June 2019, Twitter created public interest exemptions to its rules, giving the company leeway to leave tweets from government officials that violate company policies on the platform. In October 2019, the company expanded on this policy when it published a blog about its strategy for handling tweets specifically from world leaders that violate platform rules. Though Trump wasn’t called out by name in either of these posts, they were very clearly written with him in mind.
If a Tweet from a world leader does violate the Twitter Rules but there is a clear public interest value to keeping the Tweet on the service, we may place it behind a notice that provides context about the violation and allows people to click through should they wish to see the content.
Twitter did note that world leaders would still be banned from promoting terrorism, threatening violence, posting private contact information, sharing revenge porn, engaging in child sexual exploitation, or encouraging self-harm. Other than that, the rules didn’t apply, meaning that the same post that would get a regular user permanently banned from Twitter might not result in anything more than a “public interest” note or a fact-check if a “world leader” tweeted it. Twitter bent over backward to accommodate Trump’s inability to follow its rules, making the argument that Twitter as a company is somehow biased against him flat-out ludicrous.
A prime example of how Twitter pulls punches when enforcing its policies on Trump’s tweets is from May when Trump threatened anti-police brutality protesters by tweeting the phrase “when the looting starts, the shooting starts,” a line made famous by segregationist and Miami police Chief Walter Headley in 1967. While Twitter did place a warning on the tweet acknowledging that it violated the company’s policy against glorifying violence, it said it was in the public interest for the tweet to remain up. Adding that label, contrary to what MRC is trying so hard to argue, is not a form of censorship, but actually proves that Trump is held to a lower standard than other Twitter users who would likely receive a suspension.
Similarly, Facebook has rebuilt its rules of conduct around the way Trump and his campaign used the platform. Facebook has undergone a Trumpification, complete with the hiring of prominent Republican operatives, an unwillingness to enforce policies it believed would upset conservatives, and the tweaking of its newsfeed algorithm to have the specific effect of benefitting right-wing sites and harming those on the left. According to The Wall Street Journal, Mother Jones, a progressive magazine with a fairly substantial following on Facebook, was the loser in one of the social media company’s algorithm tweaks to benefit the right. Mother Jones’ traffic from Facebook tanked. Its Editor-in-Chief Clara Jeffrey tweeted that this drop in Facebook traffic likely cost the magazine $400,000 to $600,000 in annual revenue.
What was once an imperfect place to get news content has transformed into a right-wing echo chamber and radicalization ground for the dangerous QAnon and Boogaloo movements. The platform has become a cesspool in its futile quest to win the approval of conservatives. But, ironically, no matter how many policies it rolls out with the specific intention of giving Trump what he wants, Facebook still can’t escape laughable claims of “anti-conservative bias.” And the truth is that conservatives won’t stop complaining about this nonexistent bias for the simple fact that Facebook has repeatedly caved to their ever-escalating demands. This is sometimes referred to as “working the refs,” and Facebook’s response to the unending cries of bias shows just how effective it can be.
Just in recent weeks, Facebook has let Trump run ads filled with manipulated photos of Biden and allowed him promote free access to an unproven COVID-19 treatment. The platform has also sat idly by while right-wing websites game the site’s algorithm through networks of groups and pages.
The core of conservative rage about social media boils down to a belief that the rules don’t apply to them.
The conservative mindset is that only other people should have to follow rules. Take the first presidential debate as an example. The rules were clear that the audience was to wear face masks, but the president’s family didn’t. His campaign also refuses to say whether he tested negative for COVID-19 before the event. And during the debate, Trump proved himself unwilling to stick to the format and time, frequently interrupting Biden mid-response. But moderator Chris Wallace’s attempts to maintain some semblance of order quickly became a talking point in right-wing media -- a supposed evidence of his bias.
Wallace interrupted Trump more than he interrupted Biden because Trump was the far more flagrant rule breaker. Even with the home court advantage that a debate moderated by a Fox News anchor should have given him, Trump couldn’t help but play the victim. In his eyes and the eyes of his supporters, Trump was entitled to friendly treatment regardless of his own conduct.
Conservatives like Ben Shapiro and Jordan Peterson often talk about the difference between the idea of equality of opportunity and equality of outcome, stating support for the idea that everyone should have the same opportunities to succeed, but that not everyone will succeed. That same concept can be applied here. What MRC’s report says, and what Charlie Kirk’s tweet reinforces, is that anything less than a 50-50 split between the number of times Trump and Biden are flagged for breaking rules online or in debates is evidence of bias. That couldn’t be further from the truth. Actual bias is the belief that even though Trump and Biden are different people with different approaches to social media and debate, they should receive the same amount of criticism or pushback. That’s a push for equality of outcome, which isn’t really equality at all.
Even now, as this piece is being written, GOP members of Congress are in the middle of a mass freak-out because Twitter made a change to how retweets work (if you attempt to tweet an article but you haven't clicked on it yet, the platform now prompts you to consider reading it first). Yes, the change applies to all accounts and all tweets that include links, but as is always the case, any mild inconvenience they face only convinces them that they are being targeted for being conservative, delusional as it may be.