Rampant Voter Fraud in the U.S.? No, Not Likely

Adam Gravano


Like many stories, this one begins with a Donald Trump tweet. Recently, Trump tweeted: “58,000 non-citizens voted in Texas, with 95,000 non-citizens registered to vote. These numbers are just the tip of the iceberg. All over the country, especially in California, voter fraud is rampant. Must be stopped. Strong voter ID!”


Accusations of voter fraud are nothing new in the United States, although in recent years they've gained a lot more attention (one has to wonder why).


According to Google's Ngram viewer, the use of terms associated with voter fraud and electoral fraud has been on an increasing trend from at earliest 1940 to about 2008 (the furthest the viewer will let me search). While the rise might seem impressive, it's important to keep in mind that, when compared to other issues, it may not be that big. While the Ngram data is old by about a decade, looking to the 2016 election lead-up, Gallup found that four in five of those polled, yes 80 percent, were in support of some form of voter ID law. With this request for a new breed of regulation, it's important to ask if the problem to be solved by said regulation, voter fraud, is prevalent enough to justify it.


Looking to the specific type of alleged voter fraud in the tweet, we can see this variety is vanishingly small. A 2007 report titled, The Truth About Voter Fraud, by Justin Levitt of the Brennan Center, found the likelihood of non-citizen voting to be quite low as a result of high penalties (including prosecution and deportation).



In one recorded example of non-citizen registration and possible voting found by Levitt, mistakes were made in the naturalization process by groups assisting individuals in the process leading to registration, including receiving  a letter reading “Congratulations, your application for citizenship has been approved” before their swearing-in ceremony. In many cases, Levitt found that the overall effect of non-citizen voters in recent elections would have been statistically negligible (the highest percentage being .017 percent of the electorate).

Looking to a study published by The Washington Post (also by Levitt), we can see that out of 1 billion votes cast, only 31 could be tied to credible incidents of voter fraud. And looking to more recent work by the Brennan Center, one can find a cascade of work, by just about every organization or individual to have undertaken the study of the issue, stating that voter fraud is a pimple on the rump of statistical insignificance.


It's possible to list study after study and book after book, but that effort would prove tedious and duplicative. Voter identification laws are more often than not a solution in search of a problem.


One might ask exactly whether voter identification laws are guaranteed to have a disparate impact alongside the question of whether there is any sort of need for them. In 2016, the state of political science research found voter identification laws to be ineffective at suppressing minority voting, as published by German Lopez of Vox.



At worst, the discriminatory effects of voter identification laws tend to prove minimal, mainly because their effect on elections overall has proven minimal, despite the intent to push down turnout among groups more likely to vote for one party over the other. These are results that have been repeated in the political science literature: Even though the effect in the short term falls primarily on minorities and the elderly, the overall effect on all but the closest of elections is minimal at best.


So one wonders what it is that voter identification laws actually intend to do -- apart from giving the Republican Party another means of shooting itself in the foot by providing Democrats a wedge issue with which to drive turnout?


Author Bio:


Adam Gravano is a contributing writer at Highbrow Magazine.


For Highbrow Magazine

not popular
New America Media; Wikipedia Commons
Bottom Slider: 
Out Slider