Big Tech companies have continued in their efforts to silence speech they disapprove of, and the secondhand effect of Big Tech’s censorship remains as significant as ever.
As the ill effects of censorship continue to be exposed, Big Tech’s methods have become more elusive in recent years. The days of content or accounts being censored directly have dwindled. Instead, many Big Tech companies have ramped up less transparent censorship practices, such as shadowbanning, freedom of speech but not reach policies or search suppression. In other important ways some Big Tech companies have become more transparent. For example, X explains a position on a post with Community Notes — which Meta has announced it intends to emulate.
Utilizing MRC’s exclusive CensorTrack.org database, MRC Free Speech America researchers have identified and documented 983 cases of censorship in 2024, excluding cases of X Community Notes. That censorship translated to 148,076,868 times that platforms harmed social media users by preventing them from viewing content posted by the accounts they chose to follow. This phenomenon is best thought of as the “secondhand censorship effect.”
The number of times platforms harmed social media users through secondhand censorship in 2024 was a significant change from the yearly total in 2022. Researchers recorded fewer cases of censorship (517) and a greater secondhand censorship effect (275,396,336) during the midterm elections nearly two years ago.
The decrease in secondhand censorship from 2022 to 2024 reflects a drastic change in Big Tech censorship over the last couple of years. In some ways, like in the case of Community Notes, platforms have increased their transparency. But in other ways, Big Tech companies have become less transparent.
Traditional social media platforms like Facebook, Instagram and X have seemed to wind down their direct censorship of large accounts. Instead of removing huge swaths of content and suspending influential personalities, platforms are finding new ways to censor that are more difficult to identify, track and quantify — like in the case of Google search suppression. As a result, a single documented case of censorship is not always a clear indicator of the broader picture of censorship. Rather, it reflects a small dot of the picture.
In some cases, deeper research techniques have become increasingly more necessary to document censorship in an effort to capture what psychologist and censorship researcher Dr. Robert Epstein refers to as “ephemeral experiences” or subliminal unquantifiable attempts to silence speech or restrict reach.
While this makes the job of censorship researchers more difficult, it is important to stay the course. For their part, Big Tech companies like X/Twitter, after Musk purchased it, and now Meta have been responding to work that exposes their biases by publicly acknowledging the harm censorship has caused and by backing away from the most egregious forms of suppression.
Platforms
When looking at which Big Tech platforms had the largest secondhand censorship effect, several platforms are worth highlighting.
YouTube harmed users through secondhand censorship the most, as MRC researchers recorded 68,028,123 times that the Google-owned platform prevented users from accessing content they had subscribed to. This is despite the fact that YouTube had only 49 documented cases of censorship in CensorTrack for 2024.
Facebook came in second place with no fewer than 37,327,852 times the platform kept users from accessing content from the friends and pages they follow. And in 2024, CensorTrack documented 71 cases of censorship on Facebook.
Meanwhile, aside from X, Google exceeded all other platforms for the highest total number of censorship recorded in cases in CensorTrack with 434 cases. Yet, Google’s overall censorship impacted an unquantifiable number of users.
The Unquantifiable Nature of Google’s Secondhand Censorship Effect Calls for Transparency
The playing field of Big Tech censorship is changing, but MRC Free Speech America continues to stay on top of the tactics being used to silence voices that challenge the leftist narrative.
Google provides a clear example of this phenomenon. Although it was not the most aggressive censor, Google had the second-highest number of documented cases of censorship in CensorTrack aside from the X platform.
Of the 1,411 cases of censorship documented in CensorTrack in 2024 (including X’s Community Notes), 434 involved Google suppressing results or news articles presented alongside search results. The remaining 977 cases of censorship still exceeded 2022’s total of 517 cases. However, the associated harm as measured through the secondhand censorship effect went down by more than 127 million between 2022 and 2024 as Big Tech appeared to go after smaller accounts more often than in the past.
This discrepancy is made even more significant by the fact that Google’s censorship is unquantifiable. Without transparency from the platform itself, there is currently no way to determine exactly how many people conducted searches using Google’s search engine that were only presented with certain results, but the number of users impacted would likely far exceed the quantifiable secondhand censorship on other platforms. MRC Free Speech America researchers have taken an adaptable approach to such censorship tactics, identifying the ways that it can be gauged.
For example, numerous MRC Free Speech America studies have identified instances when Google suppressed search results for news articles from right-leaning media outlets. While this perpetuates an obvious harm to users seeking more than the leftist perspective on current events, this harm is not as readily quantifiable as censorship perpetrated via social media accounts.
Content Categories
It is little surprise that content related to the 2024 presidential election and campaigns topped the list of the most censored types of content. Big Tech companies kept campaign-related information from the public 47,964,105 times in the past year. An example of a censorship case that caused a widespread secondhand censorship effect was when X placed a restriction on the Trump War Room account during the first presidential debate.
Big Tech companies similarly kept election-related information from the public 33,945,996 times through the secondhand censorship effect. This included the discussion of election integrity issues. For example, Facebook placed a fact-check label on a post from Judicial Watch discussing its investigation into non-citizens voting in a primary.
There is some overlap where one censored post may have discussed both campaign or candidate issues as well as election issues and would therefore be included in both totals.
X/Twitter Censorship and Community Notes
Since X began utilizing Community Notes, MRC Free Speech America researchers have been tracking them as another form of censorship.
In 2024, X had by far the highest total number of documented cases of censorship in CensorTrack with 781 total documented cases, 428 of which were Community Notes. The collective secondhand censorship effect of this suppression amounted to 2,221,177,843 times that X harmed by hindering their ability to view alternative perspectives. Over 2 billion (2,202,995,915) of these times resulted from Community Notes alone.
These unusually large secondhand censorship effect numbers are due to the fact that MRC recorded several of the Community Notes that Elon Musk received on his posts, and his follower count is over 200,000,000.
Conclusion
As platforms continue to use more covert methods of censorship that are not as clearly quantifiable, the MRC will work tirelessly to find other ways to document and report on censorship activity. Meta’s recent announcement is a step in the right direction, but we must ensure that it lives up to its promises, and continue to press other platforms to reform their ways.
