Fact-checking Sidney Powell’s claim that 450,000 ballots in key states were marked only for Biden

[Tin tức]

Sidney Powell in an interview with Fox News:



Former federal prosecutor Sidney Powell has been waging a war against the legitimacy of the 2020 election, appearing on several conservative TV programs with unsubstantiated claims of nefarious activity that she says unfairly threw the election for President-elect Joe Biden.

On Nov. 8, Powell, who served as an attorney for former Trump administration national security adviser Michael Flynn, appeared on Maria Bartiromo’s Sunday morning Fox News show and said that Democrats are fraudulently creating ballots that exist only for Biden:

"We have identified at least 450,000 ballots in the key states that miraculously only have a mark for Joe Biden on them, and no other candidate."

The fact that many voters voted for Biden but not for candidates in other races, Powell argued, was problematic and indicative of fraud.

Bartiromo breaks down the 450,000 figure to five states, saying to Powell: "You say it’s 98,000 ballots in Pennsylvania, 80,000 to 90,000 in Georgia, another 42,000 in Arizona, 69,000 to 115,000 in Michigan and 62,000 in Wisconsin." The numbers do add up to approximately 450,000 votes and appear to have originated in a Nov. 7 tweet by the editor of a conservative news website called The National Pulse.

But this omits the context that undervoting, a practice where voters mark their choice for one race but skip making selections in other races down the ballot, is not a new or shocking phenomenon. It occurs in every election.

More than 24 million people voted in the election in the states that Bartiromo cites, making the 450,000 people who Powell says apparently voted for Biden and skipped the rest of the ballot less than 2%.
Verifying the 450,000 figure

We weren’t able to verify the exact figures Bartiromo and Powell cited. We attempted to contact Powell about how she arrived at those numbers, but did not hear back.

But using the New York Times elections data on Nov. 20, we did find that there were nearly 450,000 more votes cast for president in those five states than there were for the down-ballot U.S. House of Representatives races: 124,390 more votes in Pennsylvania; 117,458 more in Georgia; 31,123 more in Arizona; 115,717 more in Michigan; and 61,125 more in Wisconsin. We don’t have data showing that all those votes were cast only for Biden, though.

PolitiFact also reached out to top election officials in each of the states. The Michigan Secretary of State’s Office told us that the number is "easily disproven by looking at unofficial vote totals, either on our state website or individual county ones."

Arizona and Wisconsin officials both said they don’t have official statewide results yet, and are currently accepting canvass results from its respective counties.
Undervoting in American elections

There is nothing fraudulent about people voting for president or other offices at the top of the ballot and then skipping over offices further down the ballot, said Barry Burden, a political science professor and director of the Elections Research Center at the University of Wisconsin-Madison.
"This kind of ballot roll-off is entirely understandable and occurs in every election. There is nothing at all ‘miraculous’ about some voters only being interested in voting for president," Burden said. "Roll-off is most common when down ballot races are uncompetitive, but it happens even when those races are closely contested."

He pointed to Pennsylvania’s highly competitive 2016 U.S. Senate race as an example. The presidential race attracted 6.12 million votes in the battleground state, but the Senate race tallied 64,000 fewer votes.

The Federal Election Commission provides vote totals for past elections that include tables that show how many voters made selections for the Senate, the House of Representatives and, when applicable, the president.
In 2016, for example, 136,669,276 people voted for president. About 5 million fewer — 131,652,452 — voted for a member of the U.S. House.

In past presidential elections, in almost every state cited in Powell’s claim, certified elections data shows there have been several thousand more votes for president than for other races down the ballot. (An exception is Michigan in 2012, when there were more votes for the U.S. house than for president.)

Some voters aren’t given a chance to vote for a senator, but every voter does have the opportunity to vote for a member of the House. Here is a look at the undervote numbers between presidential races and House races in each state in prior presidential election years:



Robert Speel, a political science professor at Penn State University, said that 450,000 ballots marked for Biden and no other candidate sounds normal, and might even be an undercount of such ballots.

"Those who are falsely claiming potential fraud because of 450,000 ballots marked only for Biden need to report how many ballots they found marked only for Trump and no other candidate; it would likely be a similarly high number," Speel said.

Election data experts also told PolitiFact that the rate of undervotes matters more than Powell’s number in a comparison of different elections.

Charles Stewart, a political science professor at Massachusetts Institute of Technology, tweeted about this on Nov. 6. Tabulation results for early and election day voting, Stewart said, showed an undervote rate of 0.39% for both Biden and Trump.
"Thus far there's no evidence of unusual numbers of over or under votes, and indeed, it looks significantly lower than in 2016," Stewart wrote.

Powell said at least "450,000 ballots in the key states that miraculously only have a mark for Joe Biden on them, and no other candidate." We could not confirm that figure, but it does seem in line with the difference between the numbers of total votes cast in the presidential race and those cast in the down ballot House races in those states. We don’t see evidence that all of these votes were for Biden, however.

Powell presented this unremarkable undervoting statistic as evidence of fraud. Her statement fails to provide the key context that the practice of undervoting is common and does not indicate ballots have been fraudulently manufactured.

We rate this False.