Wednesday (10/28) promises to be another stressful day for Facebook, Google and Twitter.
Its top executives will be asked by senators in the United States Congress about whether they are abusing their power.
For Republicans, this is the opportunity they have been waiting for.
Two weeks ago, Twitter prevented users from posting links to an investigative report in the New York Post criticizing Joe Biden.
The social network then apologized for not explaining why it had done so before discarding a rule it had used to justify the action.
For many Republicans, this was the last straw – indisputable evidence that social media is biased and deliberately acts against conservatives.
The charge is that Silicon Valley (the region’s nickname in San Francisco, California, where several technology companies are based) is basically liberal and a poor arbiter of what is acceptable on their platforms.
In that case, Republicans like Senator Ted Cruz believed that Twitter would have acted differently if the story was about President Donald Trump.
When conservatives accuse social media companies of bias, they often refer to what they see as unjust moderation.
It is the idea that your posts are overly censored and / or suppressed.
But it is difficult to prove definitively that social networks are biased.
On the one hand, companies like Facebook and Twitter keep a lot of information under seven keys – they don’t share all of their data or reveal exactly how their algorithms work.
As a result, when Republicans complain, it is usually “anecdotal accusation (without scientific evidence)”. That is, a single example used to prove a greater trend.
They argue that Twitter “hid” a tweet from President Trump saying “when the looting begins, the shooting begins” during the protests in Minneapolis. But he did not hide a tweet from Iranian ayatollahs calling for armed resistance in Israel.
This – according to many on the right – proves Twitter’s two weights and two measures.
These examples were cited repeatedly during a Congressional hearing in July, where the heads of Google, Facebook, Apple and Amazon were questioned.
“I’ll get to the point,” said Republican Congressman Jim Jordan. “Big Tech wants to catch conservatives.”
The CEOs of the tech giants deny this charge.
But in fact, several of these companies have recently taken a much more active approach to moderation.
In doing so, they are struggling with the kinds of problems that newspaper editors face every day: what should and shouldn’t be published?
What Americans think
A poll by the Pew Research Center in August indicated that 90% of Republicans believed that social networking sites censored political views. About 59% of Democrats hold the same opinion.
So, are they right?
One of the Republicans’ criticisms of social media is that their algorithms eliminate conservative content. But that is not confirmed by Facebook data.
The data from CrowdTangle, a Facebook-owned public insight tool, brings together the most popular posts each day on this social network. On any given day, the top 10 most popular political posts are dominated by right-wing commentators like Dan Bongino and Ben Shapiro, along with posts from Fox News and President Trump.
Trump’s Facebook page has 32 million followers, almost 10 times more than his opponent in next month’s election, Joe Biden of the Democratic Party.
If the charge is that Facebook suppresses content on the right, it doesn’t seem to do it very well.
So, is right-wing content actually being favored over left-wing content?
It’s not that simple.
“I think it is a mistake to look at it as a right versus left bias,” said Siva Vaidhyanathan, a professor of media studies at the University of Virginia in the United States.
“The trend is towards content that generates strong emotions.”
He says that while “some extreme right posts” have spread widely on social media in the United States, their popularity is not proof that platforms suffer from structural bias.
“In Mexico, you can see a completely different arrangement than what is promoted,” he adds.
But if you look at what is being censored, you can see why more people on the right may have more problems in moderation than those on the left.
For example, few Democrats claim that postal voting is rigged.
President Trump and many Republicans, yes.
Facebook has a policy of labeling allegations about electoral fraud. The company argues that it is trying to combat the misinformation that can undermine faith in the US electoral system.
Even so, Republicans are disproportionately affected.
Take the Black Lives Matter movement as an example.
Facebook CEO Mark Zuckerberg has openly stated that he supports the initiative. The Black Lives Matter page on Facebook has just over 740 thousand followers.
However, another Facebook page called Blue Lives Matter has about 2.3 million followers. It aims to support the police and resist an “anti-police” narrative.
The group was criticized for appropriating the acronym BLM – and accused of racism, something the group’s founder, Christopher Berg, denies.
Berg says he believes Facebook is biased against conservative voices. Can this be true considering the popularity of the page?
“We shouldn’t look at the follower count and reach. Instead, we have to look at things behind the scenes, that individuals can impact … things like demonetizing a page,” he says.
This occurs when Facebook judges that a page has violated its rules and prevents it from making money from ads and subscriptions, for example.
Berg says he believes this is a less noticeable type of prejudice, and the right-wing pages are more susceptible.
But his suspicion is difficult to prove. Facebook does not publish a list of pages on which it has performed actions.
What about Twitter?
Twitter is a very different platform than Facebook.
Only a small minority of its users regularly publish their own content.
And a Pew study last week found that 70% of America’s highly prolific adult users were Democrats.
This makes Twitter seem like a more liberal place, but, again, it is difficult to prove that it is biased against conservatives.
What about the covid-19 pandemic?
It’s true that Twitter influenced Trump’s tweets more than Biden’s. For example, by blocking a post from Trump suggesting the flu was more dangerous than covid-19.
But at the same time, studies suggest that Trump is much more likely to spread misinformation around the disease caused by the new coronavirus.
In fact, research by Cornell University suggested that the president was the biggest driver of covid-19 misinformation.
So it may not be surprising that he is targeted by Twitter moderators “disproportionately”.
The dilemma of social media
That’s exactly why social media companies prefer not to have to moderate their platforms. Once you start deciding what can and cannot be published, political choices are made.
In fact, some Republicans treat any kind of moderation as an attack on freedom of expression.
An executive order signed by President Trump in March said, “We cannot allow a limited number of online platforms to choose the speech that Americans can access and broadcast on the Internet.”
In other words, the decision to arbitrate can be seen as philosophically anti-conservative.
Trump also said he would remove Section 230 of the Communications Decency Act. This law protects social media companies from being held responsible for the things people post – but it also protects them from the consequences of moderating or censoring conservative voices. For companies, the end of this law could devastate the sector.
And as big tech companies continue to adopt a more interventionist approach – whether against QAnon (extreme right conspiracy theory), hate speech or any other prohibited activity – accusations of bias will continue to be made.
Just as they are difficult to prove, they are also very difficult to refute.
This is where social media companies meet.
Of course, they deny acting biased.
But most Americans don’t believe them.
Have you watched our new videos on YouTube? Subscribe to our channel!