A group of prominent professors and lawyers wants Hillary Clinton to ask for a recount and forensic examination of votes in three battleground states – Wisconsin, Michigan, and Pennsylvania – because they believe they’ve found suspicious patterns that could mean the election results were hacked or rigged.
However, other leading data experts have thrown cold water on the theory, saying it’s highly unlikely.
On November 23, Green Party candidate Jill Stein announced that she was seeking a recount in the three states:
Stein raised more than $2.5 million in one day through a crowdfunding website to fund recounts.
The rumbling about the results from the three states gained more prominence November 22 when New York Magazine ran an extensive article entitled, “Experts Urge Clinton Campaign to Challenge Election Results in 3 Swing State.”
The article reported that a group of data experts and lawyers “which includes voting-rights attorney John Bonifaz and J. Alex Halderman, the director of the University of Michigan Center for Computer Security and Society, believes they’ve found persuasive evidence that results in Wisconsin, Michigan, and Pennsylvania may have been manipulated or hacked.” Halderman then wrote a detailed explanation of his theories on Medium.com.
The group’s central theory is that a foreign government could have hacked the election because of what they argue are suspicious patterns in voting results in Wisconsin (which other data experts say can be explained away by education and race variables). The Clinton campaign and various intelligence agencies have accused the Russian government of staging a series of hacks that hurt Clinton’s chances during the campaign, including the hacking of the emails of her campaign chairman, John Podesta.
The New York Magazine article, by Gabriel Sherman, says the group had a conference call with Clinton Campaign Chairman Podesta and the campaign’s top lawyer Marc Elias.
Let’s break this down step-by-step. Here’s what you need to know:
The Margins of Victory in the Three States
The margins of victory were exceptionally close in all three states, although closest in Michigan. In fact, the Michigan results were close enough that the Associated Press still has not called that state for Trump. In addition, notes Fox 6 Milwaukee, some electors have vowed to become “faithless electors” and not cast their vote for Trump when the Electoral College meets in December.
Trump’s lead in Michigan: 12,882 votes. His lead in Wisconsin: 24,081. His lead in Pennsylvania: 68,814. The leads have fluctuated slightly as votes are canvassed. Trump leads in the Electoral College 306 to 232 if you count Michigan.
Both Michigan and Wisconsin allow campaigns to request recounts after canvassing is complete. The Wisconsin Election Commission has a detailed manual spelling out the procedures. There is no automatic recount, but candidates can request recounts. In Michigan, an automatic recount is triggered only if the margin is under 2,000 votes. However, a candidate can still request a recount.
The deadlines to ask for recounts in all three states are looming. Halderman says the deadlines are Friday, November 25 in Wisconsin, Monday, November 28 in Pennsylvania, and the following Wednesday in Michigan.
Altogether, the three states amount to enough electoral votes that, had Clinton won them, she would win the Electoral College (she is already way ahead in the popular vote). See the above electoral map for how it would look if those three states flipped into Clinton’s column. She’d need all three, though.
What the Data Group is Alleging
In Wisconsin, says New York Magazine, the group argues that their analysis shows “Clinton received 7 percent fewer votes in counties that relied on electronic-voting machines compared with counties that used optical scanners and paper ballots.” They argue this might have cost her 30,000 votes.
According to The New York Daily News, “Wisconsin’s metropolitan areas, where Clinton did well, use paper ballots. Rural Wisconsin, where she did not, relies on electronic machines.” The newspaper says the group has no proof of hacking. And, of course, it’s possible that Clinton just did not appeal to rural voters.
The group has not presented similar data on Michigan and Pennsylvania, but the results in those states are close.
Professor Halderman wrote a detailed explanation of his views on Medium.com after the New York Magazine article detailing his concerns about the election results. In it, he details how a “foreign government” could hack an election. He wrote that hackers might “spread malware into voting machines in some of these states, rigging the machines to shift a few percent of the vote to favor their desired candidate.”
He is calling for a forensic examination of paper ballots and voting machines in the three states. The professor writes that states rarely check paper ballots against electronic voting machine tabulations and that the machines are largely insecure. His article contains a map in which he says that counties in Wisconsin and Michigan use optical scan ballots that could be checked against the results, and some counties in Pennsylvania do.
Ballotpedia has an explanation of the ballots used in each state.
What Other Experts Say
A series of other prominent election data analysts have said the group’s claims don’t make sense (although some of them predicted Clinton would win the election in the first place).
One of those voices is Nate Silver, of the blog FiveThirtyEight, which analyzes elections using statistics. Silver has been repeatedly posting about the claims on Twitter. He pointed out that Michigan uses paper ballots (you can see Wisconsin’s ballots here):
Michigan’s secretary of state website explains, “All voters in Michigan use optical scan ballots. Optical scan voting requires voters to either darken an oval or connect the head and tail of an arrow next to each of their choices on their ballot. Completed ballots are fed into a tabulator, which scans and records the votes.”
Silver controlled for demographics in Wisconsin and found that the difference in voting disappears. He believes that race and education levels determined whether an area would shift toward Trump.
He also didn’t find anything awry in Pennsylvania when you control for demographics.
He wasn’t the only person casting doubts on the theories. Nate Cohn of the New York Times’ UpShot page and the Cook Political Report’s Dave Wasserman were skeptical too.