Political Fraud About Voter Fraud

By Robert D. Popper,

The president’s selective statistics are red meat to supporters, but still bogus.

The Obama administration has been ramping up its rhetoric about the evil of voter identification as part of the run-up to the midterm elections. In January, Attorney General Eric Holder told MSNBC that voter fraud “simply does not exist to the extent that would warrant” voter ID laws, adding that many who favor such measures do so in order to “depress the vote.” Vice President Joe Biden claimed in February that new voter ID laws in North Carolina, Alabama and Texas were motivated by “hatred” and “zealotry.”

In an April 11 speech to Al Sharpton’s National Action Network, President Obama recited statistics purporting to show that voter fraud was extremely rare. The “real voter fraud,” he said, “is people who try to deny our rights by making bogus arguments about voter fraud.”

These arguments themselves are bogus. Consider the two studies from which Mr. Obama drew his statistics. The first, which he said “found only 10 cases of alleged in-person voter impersonation in 12 years,” is a 2012 report issued by News21, an Arizona State University project.

The project acknowledged significant gaps in its data. … Crucially, News21 noted that “nearly all the data” it received had “some vital piece of information that had been requested specifically but that was missing.” Responses lacked “important details about each case.

Given these limitations, it is hard to believe any valid conclusions about voter fraud can be drawn from this study.

Mr. Obama also cited an “analysis” showing that only 40 voters “were indicted for fraud” from 2002 to 2005. That number is drawn from an Aug. 2, 2005, Justice Department news release—which describes the department’s “Ballot Access and Voting Integrity” initiative—and from a related list of federal cases. The release mentioned 120 pending election-fraud investigations, 89 prosecutions and 52 convictions.

It is preposterous to cite that news release as proof that voter fraud is rare. The release contains no information concerning prosecutions in any of the 50 state court systems for violations of state voting laws, even though these are far more common than prosecutions for violations of federal voting laws. Even as a list of federal offenses, the news releasee is inadequate. Justice did not claim to have compiled all convictions, prosecutions or investigations—let alone all known or unsolved cases—involving federal voter fraud. The release was only a list of legal actions relating to what was then a three-year-old initiative.

More generally, judging voter fraud by counting criminal proceedings is misguided. For any crime, convictions are a fraction of prosecutions, which are a fraction of investigations, which are a fraction of known offenses, which are, in turn, a fraction of committed crimes. This is even more likely to be true of voter fraud, which is often a low enforcement priority.

We do know that this country’s decentralized system of election administration offers abundant opportunities for fraud. For example, a February 2012 report by the Pew Charitable Trusts, “Inaccurate, Costly, and Inefficient: Evidence That America’s Voter Registration System Needs an Upgrade,” found that 1.8 million deceased registrants were listed as active voters, and that 2.75 million voters had active registrations in more than one state.

Voter fraud, whatever its extent, can matter. Many elections, particularly local elections, are decided by slim margins. In January, Ohio Secretary of State Jon Husted released remarkable statistics showing that 35 local races and eight local issues were decided in the Buckeye State in 2013 by one vote or by using the state’s designated procedure, such as coin-flipping, to break a tie.

If the available evidence suggests that the amount of voter fraud is understated, the evidence that voter-ID laws suppress voting is nonexistent. In elections held after new voter-ID laws were enacted in Georgia and Tennessee, for instance, minority turnout either was stable or increased. In Tennessee, the turnout among Hispanics of voting age rose to 34.7% in 2012 from 19.2% in 2008.

When it comes to the subject of voter suppression, it is revealing that Mr. Obama avoided statistics earlier this month and relied entirely on conditional verbs: voters “could be turned away from the polls . . . may suddenly be told they can no longer vote . . . may learn that without a document like a passport or a birth certificate, they can’t register.”

The president’s speech may have been red meat for his base and good for fundraising. But it failed to engage the serious issues relating to election integrity. The coming months don’t promise an improvement.

More From The Wall Street Journal (subscription required):