The claim: “Software glitches” in Michigan & Georgia screwed up vote counts.
The facts: They didn't.
We investigated. Mistakes in two Michigan tallies were due to human error & were quickly fixed. In Georgia, software problems didn’t affect vote counts. nytimes.com/live/2020/2020…
Let's start with Antrim County, Mich.
Unofficial results had Biden up by 3,000 votes. That didn’t seem right in the Republican stronghold, so election workers checked again.
Turned out they had set up the reporting system incorrectly. They fixed it. Trump had 2,500 more votes.
In Oakland County, Mich., the winner of one local race was changed from Democrat to Republican after election workers caught and fixed a separate human error.
The votes from one city, Rochester Hills, had been mistakenly counted twice. wxyz.com/news/error-in-…
The Michigan secretary of state's office called the issues "isolated user errors" that didn't affect the actual counting of ballots and that were fixed before another layer of checks designed specifically to catch such mistakes.
But two errors, both bad for Republicans, were all the evidence many on the right needed.
The Trump bros. & others incorrectly blamed the issues on software and said they signaled wider issues with the vote. Posts & stories like these were shared hundreds of thousands of times.
Again: These weren't software issues; they were human errors that were quickly caught and fixed.
Here's the Republican clerk of Rochester Hills, Mich.: "I am disturbed that this is intentionally being mischaracterized to undermine the election.”
There were software issues in Georgia. But they had nothing to do with counting votes, according to officials & election-security experts.
In Spalding & Morgan counties, issues affected systems that check-in voters. In Gwinnett County, the issue delayed the reporting of results.
“Anyone trying to falsely connect the situations in the two states is spreading misinformation in an effort to undermine the integrity of our elections system,” Tracy Wimmer, a spokeswoman for the Michigan secretary of state, told me. nytimes.com/2020/11/09/tec…
Now the claims are centering on Dominion Voting Systems, the maker of widely used election software.
Many Republicans incorrectly tied the supposed glitches to the company. And the president even retweeted an OANN reporter who suggested Dominion was doing the Clintons' bidding.
Some problems with their claims:
Dominion was used in only two of the five counties that had issues.
And in the one county that had an issue with Dominion software - Georgia's Gwinnett County - it just delayed results. It didn't affect voting or vote counts.
So, when you hear that "glitches" undermined the vote, know that those claims contradict the available evidence.
Humans made mistakes that were quickly caught. And the few known software issues didn't affect vote counts.
Today many right-wing accounts are spreading claims that election software from Dominion Voting Systems caused widespread problems.
We looked into each claim of problems and found that they were either caused by human error or didn't affect vote counts. nytimes.com/2020/11/11/tec…
Some Republicans have pointed to voting and vote-counting issues in five counties in Michigan and Georgia.
The Dominion software was used in only two of those counties, and in every instance there was a detailed explanation for what had happened.
In the two Michigan counties that had mistakes, the inaccuracies were because of human errors, not software problems, according to Michigan officials. Only one of the two counties used Dominion.
Let's fact check the latest big voter-fraud claim: Software glitches misreported votes in Michigan.
Facts: One county caught and fixed an error in its unofficial tallies. This came even before a 2nd layer of thorough checks designed to catch such issues. freep.com/story/news/pol…
So: Yes, there was an error in the unofficial vote tally for one Michigan county.
And then county workers checked the numbers, discovered the error and fixed it. This was all before the official vote-count verification, which is more thorough & is designed to catch such issues.
We know this because of reporting by @paulegan4 at the Detroit Free Press.
State officials said the error occurred because Antrim County didn't update its election-management software.
The vote count was accurate. The problem came when the unofficial tally went to the state.
NEW: Conservative Twitter exploded yesterday with claims of proof that dead people voted in Michigan.
We looked into it. The evidence indicates that this was not fraud, but rather something much less salacious: Run-of-the-mill clerical errors. nytimes.com/2020/11/06/tec…
It began Wednesday night with tweets from @fleccas, an Ivy League offensive lineman turned right-wing internet journalist.
He had indeed found some bizarre voter files on the state of Michigan website. It appeared people born between 1900 and 1902 had sent in absentee ballots.
By Thursday morning, his posts were the talk of the Republican internet. Candace Owens, James Woods, Jack Posobiec were all sharing them, reaching millions of people. The Gateway Pundit put up a story. This was clear proof of election fraud, many said.
President Trump just tweeted again about claims of "secretly dumped ballots" for Biden in Michigan.
This is false.
These claims are based on screenshots of a mistaken unofficial tally on one site's election map that was caused by a typo that was corrected in about 30 minutes.
I know this because I just spent all day reporting it out.
I spoke to the election official in Shiawassee County who made the mistake; the election-data provider that reported it; and even the Republican consultant who tweeted the images that went viral. nytimes.com/live/2020/2020…
Here is that Republican consultant, @MattMackowiak, owning up to the fact that his screenshots did NOT show election fraud but rather an honest mistake that was quickly corrected.
“I certainly wasn’t intending to make a typo appear fraudulent,” he said.
New data shows that YouTube has cut its recommendations of fringe channels, helping reduce its spread of disinformation.
The changes have also had a knock-on effect: Fox News is now YouTube's most recommended channel alongside election-related videos. nytimes.com/2020/11/03/tec…
In several recent analyses by @gchaslot & @MarcFaddoul of recommendations on popular news videos, YouTube consistently steered people toward Fox News more than any other channel, sometimes by a wide margin.
Fox News also far outperforms other news outlets on Facebook.
Here are the top Facebook pages ranked by their share of total interactions with posts that mentioned the election, Biden or Trump over the past week. (Interactions are the only such data FB makes available.)