To question the legitimacy and fairness of recent U.S. elections is to be attacked as a probable “Russian asset” out to undermine democracy or, at best, a kook. But many voters notice something strange happening. Scholars who study elections have noticed a peculiar trend developing in the last couple of decades: Late-arriving and late-counted ballots skew Democrat blue. As Election Night drags on, the pace of Republican votes slows, and in the wee hours Democratic votes gain momentum. Political scientists call it the Big Blue Shift.
In the contentious aftermath of the last presidential election, Deen Freelon, an associate professor at the University of North Carolina, dismissed the concerns of Trump voters as falling for the kinds of claims that “trade on people’s lack of familiarity with the vote-counting process.” He said, “Things that are perfectly normal and happen in every election may look like, to the uninitiated viewer, as something irregular or problematic.”
Does that argue for leaving such matters to experts, or should it call into question why perfectly normal voting practices look problematic? Edward B. Foley, a professor at Moritz Law School at Ohio State University who coined the concept “Blue Shift,” recognizes that an election that appears unfair can be as damaging to democracy as a ballot box that has actually been stuffed.
Foley, together with MIT political science professor Charles Stewart III, authored a recent academic journal article titled, “Explaining the Blue Shift in Election Canvassing.”
“The notion of candidates harvesting extra votes during the canvass is unpalatable to begin with,” Foley and Stewart write, “but it becomes especially problematic if this harvesting process is perceived as systematically one-sided.” What if every ballot counted is legitimate and there is no fraud? The perception problem doesn’t go away. If there is the appearance that overtime ballots disproportionately add to one party’s total, it “may start to look as if, when an election goes into extra innings, one of the two teams is given extra at-bats.”
The perception problem persists in new election laws, however well-intended, whether meant to expand the number of people voting or improve election security. Beyond being perceived by the respective opposing side as opportunities to game the next election, such measures can result in consequences as far reaching, and worrisome as they are unintended.
A case in point was the Help America Vote Act, or HAVA, of 2002. The legislation passed after the fiasco of the 2000 presidential election, which turned on a few hundred contested ballots in Florida. The 2002 law required, among other things, that states upgrade voting equipment and maintain databases of registered voters. But another provision has proved to be problematic—the requirement that people who show up to vote, but who are not on the official list of eligible voters registered for that jurisdiction, be allowed to cast “provisional ballots.” According to Foley and Stewart, those new rules have “systematically” benefited Democrats in the days after elections, what Foley and Stewart call “overtime.”
They insist, however, that this benefit for Democrats is not evidence of a “nationwide conspiracy in favor of the Democratic Party, perpetuated by state and local election officials.” Instead, it is an unanticipated result of expanding the practice, not just of mail-in voting, but of accepting provisional ballots.
Provisional ballots aren’t counted until after they have been adjudicated, which may not happen for days after an election. Stewart tells RealClearInvestigations that provisional ballots are cast disproportionately by Democrats. One reason is that Democrats are less likely than Republicans to be living in the same house or apartment as during the previous election. “Democrats are more residentially mobile,” says Stewart. Younger and less affluent than Republicans, Democrats are more likely to rent their abodes. Renting doesn’t lock one into a given address the way a mortgage does. And so, Democrats move more often than Republicans, which means they are more likely to show up on election day to find they haven’t updated their addresses in the voting database. Those voters use provisional ballots, which are often not counted until election day has come and gone.
This is one reason, according to Stewart, that Democrats are more likely than Republicans to see their vote count go up in “overtime.” There are other reasons. For example, “voters with lower education (who also tend to vote Democratic) may be more likely to make errors on their registration forms,” Foley and Stewart write. More errors mean more provisional ballots, which mean delays. Rural counties tend to be Republican. Because they are sparsely populated, there are often “only one or two precincts to count,” and Stewart says that can be done more quickly than the count in the urban “places where Democrats live, which tend to process centrally, which makes for delays.”
Foley and Stewart worry that a growing gap “skewed in one direction” will lead Republicans to “regard the vote count as ‘rigged.’” But John Curiel disagrees. A research scientist at the MIT Election Data + Science Lab, Curiel tells RealClearInvestigations that “the Blue Shift isn’t the cause of distrust.” It’s just that it can be “used that way.” He blames the rhetoric of “partisan elites.”
Lonna Atkeson, Director of the LeRoy Collins Institute at Florida State University, attributes much of the Big Blue Shift to delays in counting votes cast by mail. “Voting by mail went from favoring Republicans” several decades ago, “to being equally used, to favoring Democrats” in the last 20 years. Voting by mail can create what seems like an endless overtime: California accepts ballots that arrive a week after election day. “It’s a horrible idea,” says Atkeson. “When is the election over?” She makes the case that mail-in ballots stretch out the election overtime and feed the perception that election tallies are being manipulated. “Perception is not irrelevant” to the success of a democracy, she says.
Donald Trump has been pilloried for making unproven claims that the results in battleground states such as Georgia and Pennsylvania were “rigged.” But blue shifts can be seen as having a significant role in creating the impression that the vote count wasn’t being played according to Hoyle, the famed 18th century British rules maven.
In the first hours after polls closed in Georgia, Joe Biden didn’t just trail, he fell further behind. At 11 p.m. election night, Biden had only 45 percent of the two-party vote. Which is when the blue shift kicked in. By 3 a.m., Biden had 48.9 percent of the vote. According to an analysis by Curiel and Stewart, that meant that after winning only 45 percent of the first 3.3 million votes counted, Biden won 57 percent of the next 1.4 million. “From that point forward, Biden’s vote share slowly climbed,” Curiel and Stewart write, “until he eventually received a bare majority of the two-party vote in the certified results.”
Biden also eked out a win in Pennsylvania, after late-counted votes went disproportionately for the Democrat. Is it explained by the Big Blue Shift? In one regard it is exactly the sort of outcome predicted by the blue-shift theory: Democrats overwhelmingly voted using mail-in ballots. Those took longer to count than ballots cast in person, and so votes for the Democrat were added to the count later than GOP votes.
And yet another peculiarity of the election should have muted the Big Blue Shift. Facebook CEO Mark Zuckerberg and his wife, Priscilla Chan, lavished millions of dollars in donations to state and local governments to facilitate voting and the counting of votes. For example, Pennsylvania’s Chester County received a grant of $2.5 million but still was counting votes the Friday after Election Day. The “Zuck bucks” were supposed to make the count smoother and quicker. Why didn’t the extra millions for tallying and election administration blunt the blue shift?
Political professionals may recognize this as the Big Blue Shift in action, with votes from Democratic precincts being reliable latecomers. But it doesn’t take deranged conspiracy-mongering for the average Republican voter to wonder and worry about the repeated experience of going to bed comfortably ahead only to find that just enough votes were cast for the Democrat to tip the scales at the very end.
Atkeson thinks the Big Blue Shift may be muted in November’s elections. “I don’t expect the same sort of magnitude of mail-in voting” in 2022 and 2024, Atkeson says. “The good news” is that over the last year “fewer people have been voting by mail in special elections and city elections.”
Confusingly complex ballots in which multiple precincts cut across the same counties, ill-trained or under-performing poll workers, and peripatetic voters all contribute to the Blue Shift, but that’s not all there is behind the hyperactive legal challenges to votes, and voting that have become a standard part of election day—if not election week or election month, given how long the overtime canvass lasts. A stealth provision slipped in to the infamous 2014 “Cromnibus” spending bill encourages a maximum of election litigation, creating the appearance that the winning candidate is the one with the best lawyers, not necessarily the one with the most votes.
The law in question earned its nickname by combining a Continuing Resolution, or CR, with an omnibus spending bill. It was two massive appropriations in one. The legislation not only dealt with federal spending, it became a Christmas tree, hung with a dense assortment of unrelated provisions. For example, the Cromnibus made significant changes to campaign finance law, changes that have done much to turn election overtimes into fiercely litigated contests.
It allowed individuals to contribute up to “300 percent of the amount otherwise applicable under this subparagraph.” The subparagraph in question was in 52 U.S.C. 30116, which deals with certain limits on contributions to political parties. The new provision quietly tripled what could be donated to parties for holding conventions, for buying and improving party headquarters, and for lawyers’ fees in litigating elections (or as it is put in legalese, “expenses incurred with respect to the preparation for and the conduct of election recounts and contests and other legal proceedings.”)
This has brought in millions of dollars a year for campaign lawyers of both parties. It may come as no surprise that contributing to the legislative language leading to this windfall was the Democrats’ top election lawyer, Marc Elias. (Though he built his reputation challenging election laws and litigating recounts, he is now better known for his role in passing Democratic Party monies to Fusion GPS to pay for the Steele Dossier.) RealClearInvestigations reached out to Elias to ask about the role of the Cromnibus provisions in raising money for election challenges, but he did not respond.
It is a measure of the Democrats’ focus on litigating elections that from January through June of last year, the Democratic National Committee collected $17,046,537 for its recount fund. The Republican National Committee, over the same period, raised only $4,412,362.
One lawyer who regularly litigates election issues attributes the GOP shortfall to the many lawyers without election-law expertise who brought—and lost—scores of lawsuits challenging the results of the 2020 contest. It is hard to raise money for an effort that has just failed so miserably, the attorney tells RealClearInvestigations. That said, another conservative election lawyer expects the money gap to close: “I think the RNC is well aware” that they are behind, he says.
In any case, a flush DNC can be expected to field a host of litigators to challenge vote totals and to demand the counting of questionable, late-arriving ballots. So if there is a Big Blue Shift in vote tabulation in the 2022 congressional contest or the 2024 presidential election, that prepaid and prepositioned army of Democratic lawyers would help explain it.
That presents the specter of elections being decided by lawyers, not voters. And even though it might be perfectly legal, it would create a destructive perception indeed.