Imagine a parallel universe in which the sitting president cares about holding a free and democratic election in the midst of a pandemic. Imagine that his administration is staffed with competent, incorruptible officials who devote every waking hour to stopping the virus, saving lives, rebuilding the economy and preserving democracy. Imagine that the Postal Service hires tens of thousands of extra workers to process the surge of mail-in ballots.
We don’t live anywhere near that universe. But even if we did, we’d still have to worry about the “blue shift.”
It’s a harmless-sounding term, but it describes a very real phenomenon that could trigger major disputes in vote counts across the country after Election Day, lead to weeks of litigation and, most ominously, give President Trump an excuse to challenge the legitimacy of the vote if he loses it.
The blue shift refers to the tendency of votes counted after Election Day — mostly absentee and provisional ballots — to skew in favor of the Democratic presidential candidate. This has happened in each of the past four elections, according to Edward Foley, an election law scholar at Moritz College of Law at the Ohio State University. Mr. Foley coined the term after the 2012 election, when he was trying to predict which closely contested states might become the focus of legal challenges by one or the other political party.
There were no cliffhangers that year, but Mr. Foley was intrigued by the risk, so he dug deeper into the numbers. Studying vote counts in presidential elections from 1960 through 2012, he found a clear and persistent shift toward the Democrat in later-counted, or “overtime,” votes beginning in 2004. The magnitude of the shift varied by state, but it was consistent enough to make him worry about what could happen if, say, a tipping-point state were to have a large enough shift to change the outcome of the entire election.
“Most people forget this, but Missouri in 2008 was not called for a couple of weeks,” Mr. Foley said. The initial tally on election night had John McCain leading Barack Obama by a little under 6,000 votes; as the late returns came in, Mr. McCain’s lead shrank to about 3,700. In the end, Mr. Obama did not need Missouri’s electoral votes to win the White House. “But had Missouri been a pivotal state, people would’ve been screaming over it,” Mr. Foley said.
What concerns him isn’t voting fraud, but rather how a changing vote total that tends to move in one direction can be misunderstood by an anxious public and exploited by politicians eager to preserve any advantage. “It may start to look as if, when an election goes into extra innings, one of the two teams is given extra at-bats,” Mr. Foley and Charles Stewart, a political scientist at M.I.T., wrote in a paper published this year.
This fear became reality in the 2018 midterms, when Martha McSally, the Republican candidate for one of Arizona’s Senate seats, saw her election night lead over her Democratic opponent, Kyrsten Sinema, evaporate as later-counted votes were added to the tally. Mr. Trump took note. “Electoral corruption — Call for a new Election? We must protect our Democracy!” he wrote in a tweet.
The president, who never lets an opportunity for discord go to waste, also elbowed his way into the count in Florida, where leads for the Republican candidates for governor and senator were shrinking as more votes came in. Six days after the election Mr. Trump tweeted, without evidence, that “large numbers of new ballots showed up out of nowhere, and many ballots are missing or forged. An honest vote count is no longer possible — ballots massively infected. Must go with Election Night!”
This demand has recently returned to the president’s Twitter feed. “Must know Election results on the night of the Election, not days, months, or even years later!” he tweeted last month.
In effect, Mr. Trump, clearly worried about the blue shift, has invented a new standard: The only valid vote total is the first one. Anything that subsequently alters it is suspicious, if not outright fraud. Count the days until this charge takes on a racialized tone, with insinuations of chaos and subterfuge at “urban” voting precincts, mysterious boxes of ballots suddenly discovered in inner-city warehouses and so on.
In fact, Mr. Foley explains, the blue shift is the result of a combination of innocuous factors, including the well-intentioned reforms in the wake of the 2000 election debacle that made it easier for voters to register and to cast a provisional ballot in case of any issues with their registration. Democrats have benefited more from those reforms, he says, because voter-registration problems tend to be more common among people who lean Democratic, like college students, and urban and lower-income voters.
In their paper, Mr. Foley and Mr. Stewart found that the bluer the state, the greater the shift. On the surface, that might suggest that there’s nothing to worry about: If the blue shift is most pronounced where Democrats are already likely to win, then it won’t affect the outcome. But Mr. Foley says the risk of a swing-state blue shift is greater this year. He’s focused on Pennsylvania, which Mr. Trump narrowly won by 44,000 votes in 2016, but where the Democratic candidate netted more than 20,000 votes between the initial reported total on election night and the final certified count in each of the past four presidential elections.
“It’s not huge,” he said, “but if your statewide margin is 10,000, a 20,000-vote blue shift could flip it.” Three factors could make the shift in Pennsylvania even larger this year: the pandemic, which could drive more voters to vote absentee; a move by the state to allow voters to cast an absentee ballot without an excuse; and Mr. Trump’s hostility toward voting by mail, which, supported by Attorney General William Barr, may drive Republicans to vote in person at much higher rates than Democrats.
It must be repeated: Voting fraud of any kind is rare to nonexistent. If anything, later-counted votes are examined more closely than votes cast in person. One simple way to counter disingenuous charges of fraud would be to do what a federal judge in Wisconsin ordered in April — embargo election night returns until all absentee ballots are counted. There, at least, this strategy worked. Six days passed before results were announced, no one claimed fraud, and the world kept turning.
The media have a special responsibility in this environment, Mr. Foley argues. The cultural expectation that voting results are instant, and final, on election night is a relatively recent development. Especially in a time of pandemic, major media outlets will do Americans a great service by practicing patience.
The pandemic and the crisis in the Postal Service will play central roles in this year’s vote. Together they are likely to make the blue shift bigger than usual. President Trump and his allies will, undoubtedly, exploit those changing numbers to claim fraud. If there’s any danger of fraud, however, it’s not coming from the voters.