Nearly all scholars of America’s system of locally run elections acknowledge chronic problems, including administrative incompetence, sloppy registration rolls, unreliable machinery, vote buying, and absentee-ballot fraud. But Robert Brandon, the president of the Fair Elections Legal Network and a longtime reformer, says that the current debate, “which is about people impersonating another voter, is silly.” He adds, “You can’t steal an election one person at a time. You can by stuffing ballot boxes—but voter I.D.s won’t stop that.”
Von Spakovsky offered me the names of two experts who, he said, would confirm that voter-impersonation fraud posed a significant peril: Robert Pastor, the director of the Center for Democracy and Election Management, at American University, and Larry Sabato, a political-science professor at the University of Virginia. Pastor, von Spakovsky noted, had spoken to the U.S. Commission on Civil Rights about being a victim of election fraud: voting in Georgia, he discovered that someone else had already voted under his name.
When I reached Pastor, he clarified what had happened to him. “I think they just mistakenly checked my name when my son voted—it was just a mistake.” He added, “I don’t think that voter-impersonation fraud is a serious problem.” Pastor believes that, compared with other democracies, America is “somewhere near the bottom in election administration,” and thinks that voter I.D.s make sense—but only if they are free and easily available to all, which, he points out, is not what Republican legislatures have proposed. Sabato, who supports the use of voter I.D.s under the same basic conditions, says of the voter-impersonation question, “One fraudulent vote is one too many, but my sense is that it’s relatively rare today.”
Hasen says that, while researching “The Voting Wars,” he “tried to find a single case” since 1980 when “an election outcome could plausibly have turned on voter-impersonation fraud.” He couldn’t find one. News21, an investigative-journalism group, has reported that voter impersonation at the polls is a “virtually non-existent” problem. After conducting an exhaustive analysis of election-crime prosecutions since 2000, it identified only seven convictions for impersonation fraud. None of those cases involved conspiracy.