What fact-based arguments exist that mail-in voting would lead to significantly higher instances of US voter fraud?

Question

There seems to be a consensus among prominent news sources that mail-in voting does not significantly increase the likelihood of fraud.

• The BBC: "A voter fraud database collated by Arizona State University between 2000 and 2012, found 491 cases of postal ballot fraud out of hundreds of millions of votes."
• Reuters: "Voting by mail is not new in the United States — nearly 1 in 4 voters cast 2016 presidential ballots that way. Routine methods and the decentralized nature of U.S. elections make it very hard to interfere with mailed ballots, experts say."
• New York Times: "In states that have long embraced mail voting — such as Washington State, which has been mainly using mail balloting since 2005 — those running elections see no evidence of widespread fraud."

In spite of this, President Trump and some of his allies are unconvinced:

"...they have to be very careful because you know the things with bundling and all of the things that are happening with votes by mail where thousands of votes are gathered--and I'm not gonna say which party does it--but thousands of votes are gathered and they come in and they're dumped in a location and then all of a sudden you lose elections you think you're gonna win."

--President Trump at a press conference in April

"I'm not talking about a mail-in ballot for a limited number of cases where somebody, you know, is going to be traveling around the world, and the way that the state has provided for that is, you mail in your ballot," he said. "I'm talking about a comprehensive rule where all the ballots are essentially mail-in, and there's so many occasions for fraud there that cannot be policed."

--AG Bill Barr, as quoted by CNN last month

Is it true that mail-in ballots are more difficult to authenticate or that expanding vote by mail will make it easier for individuals to commit fraud? As a US citizen who votes, I've seen first hand how polling places operate and the security measures they have in place, however I have very little knowledge of how ballots or voting machines are secured once the polls are closed. Are the alleged security issues with mail-in ballots related to the ballots themselves, the processes involved in getting the ballots to the voters, or what happens to them after they're mailed back?

I could see how there could be some issues with ballots getting sent to the wrong address or getting lost in the mail after they're filed. I could also imagine a scenario where a group of vandals goes around breaking into mailboxes and pulling out ballots but I don't see how a covert, systematic effort that's large enough in scale to change the outcome of the election could go undetected.

To clarify, I'm not looking for ulterior motives that could make one argue for or against the use of mail-in ballots. I'm simply trying to ascertain the facts on both sides.

Answer 1

The simple answer is that there are no facts to support the idea that mail-in-voting would lead to higher instances of fraud.

On the contrary, states that have experimented with mail-in-voting have not experienced an increase of fraud as reported by BBC, Brookings,Oregon Fox News, Brennan, NBS-Oregon live, Oregon Secretary of State, NPR- Washington Secretary of State

What "arguments" being floated to support the idea that fraud would increase when using mail-in-voting are speculative and imaginary.

That having been said, IMO, states that duplicate the processes used by Oregon and Washington State are likely to experience the same level of election irregularities that Oregon and Washington State report. States that choose to deviate from those processes cannot assume that their experiment will yield comparable results. (For example, Oregon processes ballots (authenticates voter's credentials on the security envelope) and scans the ballots into the tally system PRIOR to election day. Washington State processes ballots on receipt. Compare that with New York's handling of absentee ballots - Authentication of absentee ballots does not begin until election day, and tally of absentee ballots begins AFTER the polls have closed. National Association of State Legislatures

Answer 2

I think it depends on your definition. Its trivial to see that mail in ballots make it easier to do vote selling or vote coercion. The party doing the buying or coercion can watch you vote and watch you put it in the mailbox. They cannot do so in an anonymous voting booth. So, from that perspective, you can argue that yes indeed it can increase voter fraud.

As a simple theoretical approach, I point out that any safeguards you use which protect mail-in ballots can be applied to voting done in a controlled area. If there is a secure way to do mail-in, it can simply be applied to the voting booth.

The tricky thing about such claims is that voter fraud at the national level is basically a non issue. There is no evidence that national level voter fraud occured at a scale that could change an election in the history of the US. To get a sense of the scale of voter fraud, The Heritage Fountain compiled a list of known voter fraud occurances. They tabulate 1,290 cases of voter fraud in the last 20 years. Even the famously contentious race between Clinton and Trump in 2016 had a popular-vote difference of nearly 3 million votes. For voter fraud to have changed the result of that close election, it would have to be something like 10,000 times more rampant than it has ever been in the history of the United States. (and it would have to be one sided. Voter fraud on both sides would cancel each other out)

Voter fraud is a much bigger issue on small scales, such as mayoral elections, where influencing a few hundred or a few thousand people can indeed change the result of an election. At the national level, it is generally agreed that voter fraud is a boogeyman, intended to scare people into behaviors that support one's particular cause.

So we can argue that there's a reasonable path to voter fraud increasing due to mail in ballots. However, to argue that it could occur in sufficient magnitude to materially affect an election is pretty much absurd, so one should really look at "unintended" consequences, such as disenfranchisement issues. Given how small of an issue voter fraud is, a slight increase in disenfranchisement would have a materially larger effect on the democracy.

Answer 3

Where I am (germany) we have two slightly different varieties of mail voting:

1. you can have the documents mailed to you, fill in the ballot paper, put it into its envelope & close (glue) that, sign the declaration that you did this in secret and free choice and put both that form and the envelope with the ballot into another envelope and mail that e.g. to the town hall.

2. you can also go to the town hall where they have an election cabin during the time when mail voting is possible (so not only on election Sunday) and fill in everything as described above.

• Normal in-person voting happens in certain places, and everyone can go there and see that the proper procedure is followed. In addition, there must be a committee of at least n people present at any time, and the members of this committee come from different parties. Counting is also public: everyone can go and look how the votes are counted. Since this is easy to check & understand, we predominantly vote on paper (at least for the important official elections - your sports club can use election procedures that are less secure to determine their executive committee).

To pick one fraud mechanism, in general, any election procedure that is not fully secret (i.e. enforces secrecy) has a higher risk of votes being bought and sold, or voters otherwise pressed into voting a certain way: the voter can proove that they vote as promised, or detected/suspected of not voting as deemed correct by the powers-that-be. This is not possible in fully secret elections, so the buyer of the vote cannot have any guarantee that the vote is cast as they pay for - which is a serious limitation for this type of fraud, and the repressive power cannot know whom to repress.

Mail voting variety 1 is risky in that respect since there isn't anything but the signature to say that the vote was cast secretly (or indeed by the one who claims to have voted). It offers the opportunity for someone who wants to buy a vote to see that the vote is actually cast the way they want and make sure that it actually gets mailed. Or to buy the empty ballot paper and fill in as they like and forge the signature in the outer envelope.
Mail voting variety 2 is somewhat less risky in this way (e.g. you identify yourself to get the ballot paper and fill it in immediately), but it is not practically possible to get the level of supervision that the in-person voting has.

---

Not a perfect comparison since a whole lot of other things besides the mode of election changed, but: Volkskammer elections in the GDR were not fully secret: it was not mandatory to use the election cabin (and doing so was noted and had repercussions afterwards) - until the very last one, where election cabin use became mandatory making the election fully secret (also, before, no cross anywhere meant "yes" to the unified list - any "action" on the ballot (= easy to detect) meant therefore that the one who did so did not say "yes" to the existing system; and tons of further changes). In the 1986 elections, the unified list was approved with 99.74% yes votes. In the (last) 1990 elections, the PDS (successor party of the previous state party SED) got 16,4% of the votes.
Of course, not all of this drop can be attributed only to the new full secrecy of the elections: the 99+% needed a combination of lack of secrecy and actual power of the SED (ability to actually repress those who would not clearly vote as desired) and also "yes" being the default option etc. And e.g. that power was largely lost by the time the 1990 elections came up. But secret elections are part to making sure no such abuse can happen even if there is a party in power.