Twitter's Flagging of Trump's Post-Election Tweets Is Haphazard, Irrational, and Ineffectual

What is the platform accomplishing by calling further attention to the president's wild claims of voting fraud?


Illustrating again the haphazardness of Twitter's ineffectual attempts to police President Donald Trump's comments on the platform, the company has flagged a bunch of the his post-election tweets as potentially "misleading about an election or other civic process." While a private company has a right to impose whatever rules it wants on people who use its services, the rationale for these decisions is highly dubious.

Trump says things that are not true all the time, and so do many other Twitter users. Suppressing, correcting, or flagging all of those misrepresentations would be a hopeless task, and Twitter does not purport to try. But the company pays special attention to tweets that it views as "manipulating or interfering in elections or other civic processes."

Some of the rules laid out in Twitter's "civic integrity policy" seem relatively straightforward.

"We will label or remove false or misleading information about how to participate in an election or other civic process," the company says. A Twitter user would violate that prong of the policy by providing incorrect information about who can vote or how, when, or where they can do so.

"We will label or remove false or misleading information intended to intimidate or dissuade people from participating in an election or other civic process," Twitter says. Since this rule hinges on intent, it is fuzzier than the first one. The examples the company gives include "misleading claims that polling places are closed" and "misleading claims about long lines, equipment problems, or other disruptions at voting locations." On its face, the policy allows such claims when they are honestly mistaken—unverified rumors, for instance—as opposed to deliberate attempts at voter suppression.

While both of these rules are rationally related to the goal of preventing Twitter users from "manipulating or interfering in elections," the same cannot be said for the third prong of the policy, which is the one that the company invoked when it flagged the president's post-election tweets. "We will label or remove false or misleading information intended to undermine public confidence in an election or other civic process," Twitter says. That includes "disputed claims that could undermine faith in the process itself, such as unverified information about election rigging, ballot tampering, vote tallying, or certification of election results." It also includes "misleading claims about the results or outcome of a civic process," such as "claiming victory before election results have been certified."

The Trump messages that Twitter flagged certainly seem to fall into one or both of those categories. "We are up BIG, but they are trying to STEAL the Election," the president tweeted on Tuesday night. "We will never let them do it. Votes cannot be cast after the Polls are closed!" The next day he elaborated on that theme: "Last night I was leading, often solidly, in many key States, in almost all instances Democrat run & controlled. Then, one by one, they started to magically disappear as surprise ballot dumps were counted. VERY STRANGE, and the 'pollsters' got it completely & historically wrong!"

Here we have, implicitly or explicitly, "unverified information about election rigging, ballot tampering, [or] vote tallying." Another flagged Trump tweet, posted yesterday, claimed that in Michigan, where Joe Biden seems to have won by nearly 150,000 votes, "there was a large number of secretly dumped ballots as has been widely reported!" Similarly, Trump claimed "they are working hard to make up 500,000 vote advantage in Pennsylvania disappear — ASAP. Likewise, Michigan and others!"

In two flagged tweets today, Trump promised legal action against these alleged machinations. "All of the recent Biden claimed States will be legally challenged by us for Voter Fraud and State Election Fraud," he said. Plenty of proof—just check out the Media. WE WILL WIN! America First!" Later he added, "STOP THE FRAUD!"

Another three-word, all-caps message that Trump tweeted today—"STOP THE COUNT"—is unflagged so far, although the gist is similar. Likewise this tweet from yesterday: "Our lawyers have asked for 'meaningful access', but what good does that do? The damage has already been done to the integrity of our system, and to the Presidential Election itself. This is what should be discussed!"

Nor was this tweet from yesterday flagged, although it clearly implies that election fraud denied a Republican candidate his rightful victory: "Wow! It looks like Michigan has now found the ballots necessary to keep a wonderful young man, John James, out of the U.S. Senate. What a terrible thing is happening!" This Wednesday tweet likewise was not flagged, although it also intimates election fraud: "They are finding Biden votes all over the place — in Pennsylvania, Wisconsin, and Michigan. So bad for our Country!" Also unflagged: "How come every time they count Mail-In ballot dumps they are so devastating in their percentage and power of destruction?" By leaving such messages unflagged, Twitter invites users to infer that they are not "misleading about an election."

By contrast, this message from this morning was flagged: "ANY VOTE THAT CAME IN AFTER ELECTION DAY WILL NOT BE COUNTED!" As a statement of fact, that is obviously not true. But it can also be understood as a summary of what Republicans hope to achieve by legally challenging mail-in ballots that were postmarked by Election Day but did not arrive until afterward.

The problem is not just that Twitter cannot seem to apply its own rule consistently, even to its most conspicuous user. That difficulty reflects the ambiguity, subjectivity, and wide breadth of the rule, which can be read to encompass any unsubstantiated allegation of election impropriety, including claims Republicans plan to make or are already making in litigation. And that's leaving aside the difficulty of divining Trump's intent. If he truly believes what he says, he views his wild charges as legitimate grievances, which means they are not "intended to undermine public confidence in an election," let alone to "manipulat[e] or interfer[e] in elections."

The rule that Twitter is trying to apply has nothing to do with voter suppression, since all ballots already have been cast. Instead Twitter is attempting to preserve "public confidence" in the election system, a much broader goal that could justify suppressing or limiting access to all manner of potentially misleading messages, including not just allegations of fraud but also critiques of the Electoral College, speculation about the ignorance of voters, and arguments that voting is inherently immoral because it implicitly condones the unjust use of force. That's a can of worms Twitter may regret opening.

There is also the question of what practical effect Twitter's warnings have. When a tweet is flagged for violating the civic integrity policy, you have to click on "view" to see the post. Far from limiting the reach of the president's prevarications, the warning, coupled with the extra step, makes the flagged tweets seem especially worth reading, if only to see the latest ridiculous thing the president has said, which evidently was so outrageous that it prompted Twitter to intervene.

In any event, what Trump says on Twitter is also what he says in many other forums, including his weird Election Night speech, which was widely covered by news outlets notwithstanding all the outlandish claims it included. Americans, regardless of their political leanings, are understandably curious about what the president of the United States thinks, even when—especially when—what he thinks is patently absurd. What is Twitter accomplishing by delaying that knowledge for a fraction of a second?