Skip Navigation

Resources on Voter Fraud Claims

Credible research and investigation demonstrates fraud by voters at the polls is exceedingly rare.

Published: June 26, 2017

The incid­ence of voter fraud has been stud­ied numer­ous times. The consensus from cred­ible research and invest­ig­a­tion is that the rate of illegal voting is extremely rare, and the incid­ence of certain types of fraud – such as imper­son­at­ing another voter – is virtu­ally nonex­ist­ent. This page provides resources of several types:

  • Stud­ies and analyses on the preval­ence of voter imper­son­a­tion fraud
  • Stud­ies and analyses on the preval­ence of noncit­izen or nonres­id­ent voting
  • Stud­ies and analyses on fraud claims predic­ated on list match­ing meth­ods
  • Stud­ies and analyses on fraud claims predic­ated on voter list main­ten­ance issues
  • Court find­ings on the frequency of voter fraud
  • Sources on govern­ment invest­ig­a­tions aimed at uncov­er­ing voter fraud

Where avail­able, this page provides a link to an article’s full text or abstract.

Stud­ies and Analyses on the Preval­ence of Imper­son­a­tion Fraud

Stud­ies and Analyses on the Preval­ence of Noncit­izen or Nonres­id­ent Voting

Stud­ies and Analyses on Fraud Claims Predic­ated on List Match­ing Meth­ods

  • Justin Levitt, “The Truth About Voter Fraud,” Bren­nan Center for Justice, 2007. The Bren­nan Center’s report found that many repor­ted incid­ences of fraud can actu­ally be attrib­uted to “bad match­ing”: false posit­ives that arise when compar­ing the voter rolls against some other source to find alleged double or ineligible voters. These can arise from a number of sources: Errors in the under­ly­ing data can lead to an incor­rect match when one of the data sources is miss­ing inform­a­tion; Partial matches can take place when a search confuses two similar but not identical records, like two people with the same first and last name but differ­ent middle names; And insuf­fi­ciently precise match­ing criteria can lead to vulner­ab­il­it­ies like the “birth­d­ate prob­lem:” the surpris­ingly likely occur­rence that, on a large enough list, two differ­ent indi­vidu­als will have the same name and birth date.
  • Lorraine Minnite, “The Polit­ics of Voter Fraud,” Demos, 2007. Minnite found that attempts to match lists to identify illegal voting created a high number of “false posit­ives,” lead­ing to wildly inflated initial estim­ates of illegal voting. Her report high­lighted examples of this from Connecti­cut, Flor­ida, Michigan, and New Jersey.  
  • Bren­nan Center for Justice and Michael McDon­ald, “Analysis of the Septem­ber 15, 2005 Voter Fraud Report Submit­ted to the New Jersey Attor­ney General,” 2005. The report from the Bren­nan Center and McDon­ald, an elec­tion law expert at the Univer­sity of Flor­ida, found that a list of purportedly ille­git­im­ate voters was substan­tially flawed on a match of first name, last name, and date of birth alone, mismatch­ing middle initials were ignored, and other data errors occurred.

Stud­ies and Analyses on Fraud Claims Predic­ated on Voter List Main­ten­ance Issues

Court Find­ings on the Frequency of Voter Fraud

  • Texas, 2016. The Fifth Circuit, in an opin­ion find­ing that Texas’s strict photo ID law is racially discrim­in­at­ory, noted that there were “only two convic­tions for in-person voter imper­son­a­tion fraud out of 20 million votes cast in the decade” before Texas passed its law.
  • North Caro­lina, 2016. In its opin­ion strik­ing down North Caro­lin­a’s omni­bus restrict­ive elec­tion law —which included a voter ID require­ment — as purpose­fully racially discrim­in­at­ory, the Fourth Circuit noted that the state “failed to identify even a single indi­vidual who has ever been charged with commit­ting in-person voter fraud in North Caro­lina.”
  • Wiscon­sin, 2016. A federal trial court in Wiscon­sin review­ing that state’s strict photo ID law found “that imper­son­a­tion fraud — the type of fraud that voter ID is designed to prevent — is extremely rare” and “a truly isol­ated phenomenon that has not posed a signi­fic­ant threat to the integ­rity of Wiscon­sin’s elec­tions.”
  • Indi­ana, 2007. The Supreme Court, in its opin­ion in Craw­ford uphold­ing Indi­ana’s voter ID law, noted that the record in the case “contains no evid­ence of any [in-person voter imper­son­a­tion] fraud actu­ally occur­ring in Indi­ana at any time in its history.” The Court instead poin­ted to combatting the percep­tion of fraud as a basis for uphold­ing the law. Two of the jurists involved in the case at the Supreme Court and appel­late level subsequently ques­tioned the decision and the valid­ity of Voter ID laws in combatting fraud.

Sources on Govern­ment Invest­ig­a­tions Aimed at Uncov­er­ing Voter Fraud

  • Kansas, 2015–17. Secret­ary of State Kris Kobach, a long­time proponent of voter suppres­sion efforts and vice chair of the Pres­id­en­tial Advis­ory Commis­sion on Elec­tion Integ­rity, success­fully lobbied state lawmakers in 2015 to grant his office special power to prosec­ute voter fraud. He reportedly claimed to know of 100 such cases in his state. In the nearly two years since being gran­ted these powers, he has obtained nine convic­tions..
  • Texas, 2014. Texas lawmakers purpor­ted to pass its strict photo ID law to protect against voter fraud. A court filing in a lawsuit regard­ing the law stated that the Texas Special Invest­ig­a­tions Unit iden­ti­fied one convic­tion and one guilty plea regard­ing voter imper­son­a­tion in Texas from 2002 through 2014.
  • Iowa, 2014. A multi-year invest­ig­a­tion into fraud led to 27 prosec­u­tions out of 1.6 million ballots cast. In 2014, the state issued a report on the invest­ig­a­tion citing six prosec­u­tions.
  • Color­ado, 2013. A voter fraud invest­ig­a­tion, in which the Secret­ary of State alleged that 100 people voted illeg­ally, yiel­ded one convic­tion for false regis­tra­tion.
  • Flor­ida, 2012. Governor Rick Scott initi­ated an effort to remove noncit­izen regis­trants from the state’s rolls. The state’s list of 182,000 alleged noncit­izen regis­trants quickly dwindled to 198. This amended list contained many false posit­ives, such as a WWII veteran born in Brook­lyn. In the end, only 85 noncit­izen regis­trants were iden­ti­fied and one was convicted of fraud, out of a total of 12 million registered voters.
  • Wiscon­sin, 2008–11. A state task force charged 20 indi­vidu­als with a range of elec­tion-related crimes. The major­ity charged were indi­vidu­als with prior crim­inal convic­tions..
  • Maine, 2011. An invest­ig­a­tion into 200 college students accused of fraud revealed no evid­ence of wrong­do­ing. Shortly there­after, an Elec­tions Commis­sion appoin­ted by a Repub­lican secret­ary of state found “there is little or no history in Maine of voter imper­son­a­tion or iden­ti­fic­a­tion fraud.” 
  • Flor­ida, 2011. A crim­inal invest­ig­a­tion into nine indi­vidu­als who allegedly commit­ted absentee ballot fraud in a school board elec­tion led to all crim­inal charges being dismissed against all voters.
  • United States Depart­ment of Justice, 2002–05. As detailed by Lorraine Minnite in an expert report filed as part of litig­a­tion, a special­ized United States Depart­ment of Justice unit formed with the goal of find­ing instances of federal elec­tion fraud examined the 2002 and 2004 federal elec­tions, and were able to prove that 0.00000013 percent of ballots cast were fraud­u­lent. The task force released multiple docu­ments. There was no evid­ence that any of these incid­ents involved in-person imper­son­a­tion fraud. Over a five-year period, they found “no concer­ted effort to tilt the elec­tion.”