TODAY'S PAPER
49° Good Evening
49° Good Evening
OpinionCommentary

Arguments that mail-in balloting will undermine election are wrong

Workers prepare absentee ballots for mailing at the

Workers prepare absentee ballots for mailing at the Wake County Board of Elections in Raleigh, N.C. on Sept. 3, 2020. Credit: AP/Gerry Broome

There is a line of rhetoric about the upcoming presidential election that has, by now, set like concrete. The results will be highly dubious, the argument goes, because of the broad expansion of mail-in ballots. The certainty that usually accompanies an election will be out the window as we await vote-counting and, more conspiratorially, as fake ballots are injected to shift the result. It will be a maelstrom of confusion and cheating that will damage the country.

It's also almost entirely nonsense, generated from a combination of misunderstanding, old false arguments and new political motivations.

Why? Let's assess each point independently.

- - -

Few states are sending ballots without requests

A centerpiece of President Donald Trump's rhetoric on this point is his assertion that states are sending out ballots willy-nilly to anyone who has ever filled out a voter registration form.

"This is being done on purpose. They know it's no good," Trump said last week. "They know it's — it's going to be fraudulent. It's going to be fraud all over the place. Who's getting the ballots? Who's sending the ballots? They have people saying you don't need a verified signature. This is a serious threat to our democracy, and the Democrats know that."

He often contrasts this strategy of ensuring that voters have ballots to complete with a more traditional absentee ballot process, something he himself uses and has publicly advocated for.

That's also exactly what most states are doing.

As of Thursday, there are 10 places in which voters will be sent an actual ballot by mail, the thing Trump rails against. Five did so before the coronavirus pandemic: Colorado, Hawaii, Oregon, Utah and Washington. Five began doing so this year: California, the District of Columbia, Nevada, New Jersey and Vermont.

You've probably noticed that there aren't many swing states among those 10. Only Nevada is a state in which the contest might be close and which is sending ballots. Nevada is also the focus of Trump's "you don't need a verified signature" assertion, which we'll get to in a bit, but it's worth noting that this is not true.

Among the seven states that were closest in 2016, two besides Nevada made changes. New Hampshire introduced voting by mail, though with an application. Wisconsin will send absentee ballot applications to every voter.

In total, 51 million voters will receive a ballot in the mail, according to a Washington Post analysis. About half of them are in California and Washington state, both places that Trump had no shot of winning.

- - -

There's almost no fraud in American elections and no reason to think this year will be different

It has been fascinating to watch how Trump's focus on fraud has shifted over the past four years. Before 2016, he warned that fraudulent voting would occur at polling places in Pennsylvania. Then, when he won Pennsylvania, he began claiming that his popular-vote loss was due to the commission of fraud at a truly staggering scale, the sort of thing that might seem detectable. Then, as states began expanding voting by mail, given the pandemic, he turned his attention to those mail ballots.

In no case has Trump's rhetoric matched reality. There's no more significant threat of rampant voter fraud from those mail ballots than there was from people voting multiple times in Philadelphia or from millions of noncitizens voting in California — all claims he has made. Each has been specifically tailored not to observed reality but to the political point he wants to make.

Trump and his allies often will allege widespread voter fraud by pointing to things that might, in theory, be used to commit fraud. A lot has been made, for example, of people encountering batches of ballots intended for voters who'd moved or died. But there is a wide gap between "having the ability to cast a ballot fraudulently" and "casting a ballot fraudulently," a gap that necessitates, among other things, violating state or federal law. There's a wider gap still between casting a ballot and "having that ballot be accepted as legitimate."

The amount of fraud in recent American elections has been de minimis. Sure, dubious things happened in 1960, but this isn't 1960. And, sure, dubious things happen, but rarely at any significant scale and at no recent point at a national one.

You will hear assertions to the contrary. You will hear, for example, that former president Jimmy Carter co-chaired a commission that determined that mail-in voting was risky. In fact, the commission wrote that voting by mail was "fraught with risk," a determination based on a report from another organization. It was also 15 years ago, before mail-in voting was broadly expanded and additional checks were added to the process. A Post analysis looked at nearly 15 million votes cast by mail in 2016 and 2018, and it found fewer than 400 possible cases of double voting or a vote cast for a dead person. If each of those cases was, in fact, intentional fraud — which has not been determined — it's equivalent to having $10,000 in the bank and seeing someone purloin a quarter. (It's also worth pointing out that Carter endorsed voting by mail this year.)

That qualifier that fraud has not been proved in those 400-odd cases is important because it's common for sweeping allegations of fraud to be made that collapse upon further scrutiny. A report from Texas last year, for example, claimed that tens of thousands of noncitizens had voted in the state, a claim Trump quickly elevated. But it was not true, ignoring that many of those voters had become citizens before they voted. A review subsequently found only 80 instances in which people might potentially have cast improper votes.

There have been situations in which systematic fraud has occurred. In 2018, a campaign consultant allegedly coordinated a system of collecting and filling out ballots on behalf of voters in North Carolina. The election results were thrown out, despite the level of fraud not being sufficient to affect the results. It was a Republican campaign for which the alleged fraud was committed; a Republican ended up winning the new election.

Trump talks less about that example than Paterson, N.J. There, evidence of fraud was detected in a local election and charges filed against both a (Democratic) candidate and two campaign workers.

An important point in both of those cases is that the fraud was detected. Efforts to affect an election are rare, but, in two cases in which people tried it, it was uncovered in short order. Trump has repeatedly suggested that the number of rejected ballots in Paterson is itself a signifier of fraud, but the point is entirely that efforts to commit fraud have to overcome robust checks designed to detect it. Local officials may, in fact, be overzealous about tossing ballots out of concerns of fraud, something that promises to be a focus of legal fights after this year's contest.

Those checks are a reason that some of Trump's wilder scenarios, like an election flooded with ballots sent by foreign actors, don't make much sense. How, for example, are (say) Russian agents going to slip millions of ballots into the mail undetected? More critically, how are they going to manage to successfully mimic voter signatures in a way that passes detection upon scrutiny? (A forgery expert with whom The Post spoke called this scenario "highly unlikely.")

Were this the sort of election that takes place in a fourth-grade classroom, where kids check boxes on a slip of paper and drop it in a ballot box, fraud would be easy. But, happily, the nation's process for voting is a bit more sophisticated than that — and the dearth of demonstrated fraud reinforces that.

It's worth noting, too, that Trump's claims of fraud are vague enough to be inconsistent. At a rally in Pennsylvania this week, he reverted to his 2016 allegations that people would cast ballots illegally at polling places.

"If you see somebody cheating, you've got to turn them in," Trump said. "And you probably will. It's going to be a mess."

Actually, you almost certainly won't. But Trump would probably be content with his supporters accosting those who they thought shouldn't be voting on Nov. 3.

- - -

It has been common in recent years for ballots counted after Election Day to favor Democrats — for predictable reasons

The prospect of vast numbers of mail-in ballots arriving on — or, in some states, after — Election Day has prompted Trump and his allies to suggest that this is part of an effort to throw the election to his opponents. Again, this is almost entirely predicated on the idea that those votes are subject to rampant fraud, something that is itself unfounded.

But it's theoretically bolstered by the demonstrable fact that votes counted after Election Day tend to favor Democrats. If your theory is that Democrats cheat and will cheat using mail-in ballots, a Democrat-friendly post-election tally seems to reinforce your point.

In reality, though, this has been a common phenomenon in recent years. In 2016, for example, Trump and Hillary Clinton were running about even in the popular vote on Election Day. It was only the post-election tally, particularly in California, that gave Clinton her broad popular-vote lead.

Generally speaking, there is a number of reasons that this vote tends to be more Democratic. One is that more voters means more ballots to count, and big, urban counties tend to have more voters — and more Democrats. Another is that voters who vote less frequently (meaning that they are more likely to require a provisional ballot) or those who register to vote on Election Day tend to align with Democratic constituencies: non-White voters and younger voters in particular.

This year, though, that partisan gap will be much broader. Democrats have been consistently more likely than Republicans to say they plan to vote by mail-in polling. On Wednesday, for example, Quinnipiac University released a poll showing that about a third of the electorate plans to vote by mail; that group prefers former vice president Joe Biden over Trump by 42 points. Among the half of respondents who plan to vote in person on Election Day, Trump leads by 22 points.

The result is what has been called the "blue shift" — results moving more in the Democrats' favor after Election Day. Trump has repeatedly suggested that this is some nefarious change, as he did when California's votes in 2018 shifted control of the House to the Democrats by a wider margin or as he did when Florida's close Senate race couldn't be called for the Republican candidate immediately. He should have learned about the "blue shift" in 2012, though, after his wild assertions on election night, when he mistakenly thought that Mitt Romney had beaten President Barack Obama in the popular vote.

What this means, though, is that the results of the election as they stand on Nov. 3 will almost certainly not reflect the actual vote.

This isn't a function of fraud and scheming. It's a function of counting votes, and a predictable one.

- - -

This may be a problem that could have been ameliorated

The reason that Democrats are so much more likely to vote by mail is simple: They are much more likely to express concern about contracting the coronavirus. The effort to shift states to expand voting by mail wasn't a function of building systems to commit fraud, but to give voters the opportunity to vote while still maintaining a safe distance from other people.

Many of these efforts to foster more mail-in voting were initiated earlier in the year, when medical and political experts were scrambling to determine how best to conduct an election in the midst of a highly contagious virus. Over time, we've learned how we might effectively manage an in-person election, with even the country's top infectious-disease expert, Anthony Fauci, endorsing the idea if sufficient protective measures are used.

Voting in person remains riskier than voting from home, obviously, but assuming the use of masks and social distancing, the risk of voting at a polling place is relatively low. This is particularly true if early voting is available, allowing voters to better avoid interacting with other people.

It's likely, though, that this ship has sailed. We can expect a combination of factors that will lead to more votes being cast by mail and counted in the days after the election and that those votes will be more heavily Democratic. We can further expect that this will be cited as evidence of fraud by a president eager to prevent those votes from being tallied.

Everything written above, though, should serve as evidence that such assertions are both false and politically motivated.

Philip Bump is a writer with The Washington Post.

Columns