Did Monitors Discover a 9,626 Vote Error in the DeKalb County, Georgia Recount?

On November 18 2020, Georgia Republican Party Chair David Shafer tweeted that a “9,626 vote error” had been unearthed in DeKalb County, Georgia — a claim which, naturally, spread rapidly to other platforms:

One of our monitors discovered a 9,626 vote error in the DeKalb County hand count. One batch was labeled 10,707 for Biden and 13 for Trump – an improbable margin even by DeKalb standards. The actual count for the batch was 1,081 for Biden and 13 for Trump.

— David Shafer (@DavidShafer) November 18, 2020

Shafer’s tweet claimed:

One of our monitors discovered a 9,626 vote error in the DeKalb County hand count. One batch was labeled 10,707 for Biden and 13 for Trump – an improbable margin even by DeKalb standards. The actual count for the batch was 1,081 for Biden and 13 for Trump.

Headlines about a “9,626 vote error in DeKalb County” immediately spread to the usual disinformation and hyperpartisan sites, tagged as a “report” and framed in flexible language. Shafer’s tweet appeared to be the sole source for articles reporting on that specific number of votes:

A monitor has reportedly discovered a 9,626-vote error in Georgia’s DeKalb County’s hand recount, according to Georgia Republican Party chairman David Shafer.

“One of our monitors discovered a 9,626 vote error in the DeKalb County hand count. One batch was labeled 10,707 for Biden and 13 for Trump – an improbable margin even by DeKalb standards,” Shafer said. “The actual count for the batch was 1,081 for Biden and 13 for Trump.”

According to Shafer, the error, if not discovered, would have essentially given Biden enough votes to “cancel out Trump’s gains from Fayette, Floyd and Walton.”

Social media users inevitably read into the claim that the purportedly “found” ballots netted Donald Trump 9,626 votes:

So Biden lost 9626 votes?

— Brian Norce (@blnorce) November 18, 2020

In a tweet with far less engagement than Shafer’s tweet, political analyst Dave Wasserman addressed the rumors. Describing Shafer’s claim as “grossly misleading,” Wasserman included crucial context — that the “recording error” during the DeKalb County hand count did not affect the outcome of the election:

This is grossly misleading. The recording error during the DeKalb hand count doesn’t impact the overall GA margin, because the tally was recorded correctly in the original count. But of course, this post has 8k RTs and 17k likes. https://t.co/q7ctYUwPmL

— Dave Wasserman (@Redistrict) November 18, 2020

Georgia’s voting implementations manager Gabriel Sterling answered a question about Shafer’s allegations in a video clip uploaded by Georgia’s WXIA-TV shortly after the tweet went viral. Sterling described Shafer’s claim as hinging on a “non-issue” and a “non-event,” apparently having addressed the 9,626 found votes claim previously in the call.

He confirmed the interviewer’s statement that the 9,626 votes were included in the initial machine tally. Sterling appeared to spend much of November 18 2020 running interference thanks to Shafer’s claim, and another election official spoke to a local news organization to further explain the purported discrepancy:

According to Sterling, the audit is running its course, and quality control is in place across Georgia’s 159 counties.

Releasing internal audit results, like Shafer’s tweet, is premature.

“This is why you don’t release interim audit results because human beings are going to make mistakes in a hand count audit. The error was discovered, it was corrected,” Sterling said.

The error will not change the outcome of the election in Georgia. Biden is the projected winner of the Peach State.

Baoky Vu, a member of the DeKalb County Board of Registrations and Elections, blasted Shafer’s Twitter statement.

“It’s a childish and irresponsible tweet from David Shafer, who I know,” he told Decaturish. “We are coming out with the evidence to refute the implication that somehow DeKalb undercounted votes for Donald Trump. Obviously, the President of the United States has done his part to lower the standards of leadership and facts. Sadly, his sycophants won’t be nominated for Profiles in Courage awards anytime soon.”

Sterling said the error would have been discovered and corrected, even without the monitor, because audit consultants use quality control measures.

Both Sterling and Vu suggested Shafer’s claims were intentionally misleading, intended to “stir the pot” by advancing the idea that a partial count somehow represented the final outcome.

Shafer’s Twitter claim that a monitor “found” 9,626 votes in DeKalb County was quickly interpreted as a claim that auditors had unearthed 9,626 uncounted votes for Donald Trump. In reality, those votes were present in the initial “machine count,” and that number was only relevant in the context of the recount. Finally, even Shafer did not claim all 9,626 votes were cast for Trump.