I posted the following on December 1, 2016. It seems timely today. The bottom line: voter fraud is very unlikely to have swung the popular vote in favor of Hillary Clinton, but it is all too common.

Democrats have long asserted that voter fraud is rare. Recently, we heard from them that questioning the results of an election would “undermine democracy”. In fact, voter fraud is routinely characterized by the left as “fake news“, and even worse, as a racist narrative! How convenient. But in the wake of the Donald Trump victory, we’ve been hearing about electronic voter fraud from the same crowd that’s been imagining Ruskiis under their beds for months (to steal a phrase from Glenn Reynolds). Fear not: voting machines are not connected to the internet!

This week, however, Donald Trump stirred the pot once again by tweeting that he would have won the popular vote if not for the “millions” of illegal votes for Hillary Clinton. Hilarity ensued, and not only on the left. All the pundits say that Trump has no data to support his claim. He probably never looked for it, and he probably doesn’t care. As Ed Driscoll notes at Instapundit, perhaps “stray voltage” is simply part of his plan.

Trump’s claim really does sound outrageous, but a review of the recent history of actual and potential election fraud shows that it might not be as radically far-fetched as we’ve been told. DiscovertheNetworks.org (DTN) provides a three-part compilation of voter fraud research and cases spanning the last 30 years. Pertinent detail on each case or finding is provided, and each item is sourced. The cases span the country and include fraudulent voter registration efforts, dead and ineligible voters (including pets) on the rolls, multiple registrations across jurisdictions, homeless voters casting multiple votes, fraudulent absentee ballots, vote buying, voter impersonation, and failure to provide absentee ballots to deployed military personnel. ACORN, by the way, is well-represented on the list.

Many of the cases on DTN’s list involve anywhere from a handful of fraudulent votes to several hundred. Of course, it’s likely that only a small percentage of fraudulent votes are ever detected. But there are cases on the list of fraudulent registrations numbering in the thousands, and counts of ineligible voters appearing on voter rolls numbering in the hundreds of thousands and even millions.

One of the studies cited by DTN was commissioned by The Pew Center on the States, published in 2012. It found that there were 24 million invalid or “significantly inaccurate” voter registrations in the U.S. And just before every election, said the report, election officials are inundated with a flood of new and often questionable registrations.

Another study cited by DTN appeared in the journal Electoral Studies in 2014. It said “… based upon extrapolations from the portion of the sample with a verified vote … 6.4 percent of non-citizens voted in 2008 ….” The authors admit that there are reasons to think 6.4% is an under-estimate. That’s especially true given the focus on immigration policy in this year’s presidential campaign. But if that percentage was repeated in this year’s election, and given 24 million non-citizen residents in the U.S. (legal and illegal), then roughly 1.4 million non-citizen votes would be included to the 2016 popular vote total. The researchers acknowledge that this group tends to vote heavily for democrats. The overlap between these votes and those arising from the other kinds of voter fraud by Pew is certainly not complete, so the fraudulent vote total is likely to be well north of 1.4 million.

The electoral college was designed to discourage voter fraud in states dominated by a single party. Vote margins beyond a simple majority provide no incremental reward in the electoral college, the reasoning goes. That doesn’t mean election fraud doesn’t occur in those states or that it isn’t motivated in part by presidential politics. Moreover, state and local races can still be contested in so-called “one-party” states and may be subject to manipulative efforts. In such cases, presidential votes might well ride on the coattails of candidates for state and local offices.

The recent tide of republican success in congressional races and at the state level does not suggest that election fraud is benefitting democrats in more highly contested states. Perhaps it goes the other way or is roughly balanced between the parties in those states. But most people who believe Trump’s tweet would probably say that fraud must be concentrated in heavily “blue” states like California and New York. If so, it would be unbalanced fraud.

The magnitude of voter fraud in the presidential election is plausibly in the range of 1 – 2 million and it could be even higher based on the research and other information cited above. That total, however, is split between the parties. For the sake of argument, if 2 million fraudulent ballots are cast and republicans garner 30%, or 600,000 fraudulent votes, then the contribution to the democrat vote margin is just 800,000 (1,400,000 – 600,000). Hillary Clinton’s popular vote margin was 2.9 million (less than the margin in California alone). Given that total, Trump’s claim is a real stretch, but his “guess” at the number of fraudulent votes is probably well within an order of magnitude. That might be surprising to some detractors.

What should be obvious is that voter fraud is a major problem in the U.S., and it undoubtedly swings some races at state and local levels. I have been lukewarm with respect to voter ID laws, but I am persuaded that they are a necessary step in the quest for electoral integrity. (Whether IDs must be government-issued is a separate matter.) The argument that these laws are discriminatory is true to the extent that we wish to prevent ineligible individuals from voting. That’s a good thing. The argument that it is racist is sheer stupidity: citizenship should bring privileges. That is not a position on immigration policy. Voter ID laws place a simple burden on citizens to prove that they are legitimately entitled to full participation in the democratic process. If you can’t be troubled to identify yourself, you should expect multiple obstacles to sharing in the fruits of modern society.

Postscript: I just ran across this post, which makes some of the same points I’ve discussed above, but it says that there are roughly 20 million adult non-citizens in the U.S. today.