After the 2016 general election, a large number of elected officials - Republican primarily - expressed an interest in checking into voter fraud, a subject the incoming president said he planned to pursue. One of them was the secretary of state in Oregon, Dennis Richardson. In September 2017, having reviewed voting cases around the state, in an election in which two million votes were cast, he said his office had identified 56 voter fraud cases cases to the state’s attorney general’s office.
In April 2019, after reviewing the cases, that office said it had obtained guilty pleas in 10 cases, involving people registered in three parties (including both Democratic and Republican) and none at all. The 10 were charged with felonies initially, but those charges were negotiated down to lesser offenses. (Some of the others remain formally under investigation, though most simply never led to charges.)
Most of the cases had to do with voting both in Oregon and Washington state. One person was “was suffering from kidney infections which impacted his cognition” - he was confused. Another was a woman who filled out a ballot for an elderly parent in Washington, and then her own proper ballot back in Oregon. Another was an Oregon student living in Colorado, and forgot that he had already voted in one state when he cast a ballot in another.
All of those, and the others, were offenses - these were voters who broke the law, and shouldn’t have, and they’ve been fined. But none of them represent any kind of grand conspiracy, and their numbers were so very low - 10 voters, ultimately possibly a few more - out of two million votes cast, that they do not represent a significant threat to the election regime in Oregon.
That’s more or less the norm in the United States: A few people who screw up, but hardly any cases of messing with the system in such a way as to try to influence the outcome of elections.
It’s not that voter fraud, of one sort or another, doesn’t or hasn’t ever occurred. The Robert Caro volume Means of Ascent about Lyndon Johnson contains a few hundred pages about how Johnson stole - not too strong a word - the 1948 Democratic Senate primary in Texas, through all sorts of means including the mass fabricating of votes. It is one of the most extreme cases of voter fraud - though ordinary voters generally had little to do with it - in American history.
The most striking recent case was in North Carolina, a well-publicized case in that state’s 9th congressional district, aimed specifically at electing one candidate, Republican Mark Harris. Months of investigations followed, along with an eventual order to hold a new election. During one hearing, a news report said, “State investigators established their theory of the case — that a Republican operative, Leslie McCrae Dowless, directed a coordinated scheme to unlawfully collect, falsely witness, and otherwise tamper with absentee ballots — and workers who say they had assisted him in the scheme delivered damning testimony describing their activities.”[https://www.vox.com/policy-and-politics/2019/2/21/18231981/north-carolina-election-fraud-new-nc-9-election] That case involved hundreds of fraudulent votes, evidently enough to change the outcome of the election.
But both of those cases (historical and recent), and all of the other small numbers of organized vote-tampering cases, involve political activists (sometimes though not always involving candidates) who subvert the actions of the voting public. The voting public is not the problem - not more than a flyspeck, at least.
“Voter fraud”, then, seems a misnumber. “Political activist fraud”, as uncommon as it, too, is, may be a more appropriate way to express what’s going on.