By Jon Cassidy | Ohio Watchdog

We made this fake payroll stub in 10 minutes with Excel. You could do better.

Ohio’s voter rolls are swollen by hundreds of thousands of non-voters, presenting an opportunity for massive elections fraud, a Watchdog analysis reveals.

Here’s how the fraud would work, why it’s possible, and what can be done about it.

HOW THE SYSTEM WORKS NOW

Last year, a bill to require photo identification at the polls cleared the state House but died in the Senate.

Critics said the photo ID proposal was unnecessary.

“We have never seen an instance of voter impersonation,” said Peg Rosenfield of the League of Women Voters. “We have never even heard of one.”

FRAUD KIT FOR BEGINNERS: HERE’S WHAT YOU’LL NEED TO GET STARTED

While actual voter fraud is indeed rare, anyone determined to defraud the Ohio system would need just two things to carry that out:

* A database of Ohio residents who are registered but unlikely to vote.

* One of the 12 types of ID that Ohio law requires — including utility bills, bank statements, school transcripts, and paystubs — as long as the document has an address that matches the registration record of that unlikely voter.

On Election Day, the fraudsters and their accomplices could simply show up with forged documents indicating they are one of Ohio’s many unlikely voters — and cast a ballot.

Impossible? That’s the basis of an alleged fraud carried out in Providence, Rhode Island. Rep. David Cicilline, D – RI, faced accusations by his primary opponent this year, Anthony Gemma, of repeatedly organizing fraudulent voting over the last decade, including in-person fraud. Gemma had one on-the-record witness, and transcripts purported to be from other witnesses. Witnesses said conspirators impersonated registered voters who were determined unlikely to turn up on Election Day.

Now, Ohio polling stations have a “signature pollbook” that allows poll workers to compare voter signatures to the ones on file. If one official notices a discrepancy and all four agree that the signatures don’t match, you get a provisional ballot instead of a regular one.

That’s what stands between the electoral system and fraud.

HOW TO DEFRAUD THE OHIO SYSTEM

To determine whether such a scam would be possible in Ohio, we asked Earl F. Glynn of Watchdog Labs (our sister organization) to create a database of unlikely voters. He returned with a list of 306,152 registered voters who won’t be showing up at the polls in November. It’s sorted by county and precinct, and includes addresses.

We say they “won’t be showing up” to vote because they apparently haven’t in more than a decade. They are voters who registered in 2006 or earlier, but who haven’t cast a vote in the 12 years the state has been tracking that data.

Using that database, the fraudsters could easily fabricate phony documents with the unlikely voters’ names and addresses. Free paystub templates and payroll software abound online. Counterfeit utility bills are as easy as a Google search, or pasting a new name and address onto a real one and making a photocopy.

The only part that requires a bit of know-how is writing code to sort through massive voter data files published by the Secretary of State, and Glynn’s work shows that’s not much of an obstacle for someone who knows databases.

We asked Glynn what skills were needed for the job.

“Exploratory data analysis and computer programming skills are needed to connect the dots,” he said. “I try to use the best software tool for a job, but the size of the Ohio voter file created some technical problems to overcome, especially in scrutinizing the voter history data.  The free R software tool is great for analysis but has problems with larger datasets, such as Ohio’s voter file.  A database tool like MySQL can easily store the data for Ohio’s 7.8 million voters, but is not as flexible with analysis.”

We’re not advocating actual vote fraud, of course – just demonstrating that the old, practical barriers to vote fraud are no match for technology, and that bloated voter rolls pose a hazard.

PHANTOM VOTERS

In all, Glynn found 1.3 million registered voters – out of 7.8 million total statewide – who have never cast a vote.

“About 500,000 of the 1.3 million who have no history of casting a ballot registered to vote in 2011 or 2012, but a number of voters with or without voting history have not voted in years and are bloat in the voter registration file,” Glynn writes.

According to the United States Election Assistance Commission, most elections officials nationwide remove voters from active rosters after they miss two or more federal elections in a four-year period.

We settled on three elections – 2006, 2008, and 2010. The 306,000 non-voters above have missed all of those elections.

In addition to those folks, who have never voted, there are another 440,000 or so voters who voted long ago, but haven’t voted since 2006, Glynn found. Many of them have died or moved away.

A list that included those voters would be twice the size of Barack Obama’s margin of victory over John McCain in 2008.

Bloated voter rolls provide opportunity for fraud. Indeed, Judicial Watch and True the Vote sued Republican Secretary of State Jon Husted last month, alleging that Ohio has failed to trim obsolete records. Their lawsuit claims that typical voter rolls cover around 71 percent of the voting age population. Ohio’s statewide average is closer to 88 percent, Glynn found.

He also found that the quality of voter list management varies from county to county. (See the numbers.) His other findings point to real potential for fraud:

• In Wood and Lawrence counties, there are somehow more registered voters than people old enough to vote.
• For most of the last decade, Cuyahoga County had voter registration higher than 100 percent of its voting age population, reaching 110 percent by the 2008 election. Officials finally got around to trimming the rolls after that, removing more than 210,000 of the 1.1 million names then on the county’s rolls.

Officials have purged nearly half a million names from voter rolls since the last election, but there’s still plenty of room for abuse.

Husted recently said that he expects “the General Assembly will take up a more strict version” of voter ID requirements next year, eliminating some of the 12 types of ID now accepted.

“We need to streamline that because it’s really hard for a poll worker to know exactly what they’re supposed to be checking and I’m quite confident the legislature is going to take that issue up,” Husted said.

NOTHING TO SEE HERE, FOLKS

Critics of photo ID laws say that sort of argument is a pretext for disenfranchising voters that tend to support Democrats. They’re right to point out that vote fraud is rare, in this country, at least.

News21, an investigative reporting project from Arizona State University, found 2,068 cases nationwide since 2000 where vote fraud was alleged.

With 77 cases, Ohio came in No. 7 in the nation, trailing Georgia, Washington, Kentucky, Kansas, Connecticut, and California.

Only a handful involved in-person voter impersonation.

Kevin Drum writes in Mother Jones that “a moment’s thought suggests that this is vanishingly unlikely to be a severe problem, since there are few individuals willing to risk a felony charge merely to cast one extra vote and few organizations willing or able to organize large-scale in-person fraud and keep it a secret.”

But fraud doesn’t have to be large scale. Even small agencies control multi-million dollar budgets. If a union is willing to set up a councilman for a drunken driving charge, what would stop it from padding a few thousand votes?

DOES VOTER ID ‘DISENFRANCHISE’ LEGIT VOTERS?

If there isn’t much evidence of actual fraud, there’s even less empirical support for the idea that photo ID laws lead to disenfranchisement. Drum, of Mother Jones, cites two peer-reviewed studies that conclude, as one put it, “voter-ID laws appear to have little to no main effects on turnout.”

Most of the media noise to the contrary – that the laws discriminate against minorities and the elderly, that 11 percent of citizens have no ID – can be traced back to a single study by the Brennan Center for Justice at New York University, based on a 2006 phone survey of 987 self-professed American citizens.

The researchers asked people if they were Americans and if they could get the papers to prove it by “tomorrow,” and then extrapolated from the survey what that meant for the country as a whole.

To point out just one obvious problem, the survey weighted responses from minorities, because they used to be underrepresented in phone surveys. The study’s footnotes refer to a 1990 study showing that 13 percent of black households had no telephone compared to 4 percent of white households. By the time of the Brennan Center survey in 2006, cell phones had made those landline numbers obsolete, with people of all sorts dropping landline service for cell.

How heavily did the Brennan Center put its thumbs on the scale? We don’t know, because the published study doesn’t include the raw numbers, nor does a defense of the study published five years later.

Another problem: many respondents had no idea what they were talking about; one in eight professed to have both nationalization papers and a U.S. birth certificate, an impossibility.

A more fundamental problem, identified by The Heritage Foundation, is that the study made no distinction between people who profess to be American citizens over the phone and actual American citizens eligible to vote, or the even smaller set of actual registered voters.

Its authors found the Brennan center study “dubious in its methodology and results and suspect in its sweeping conclusions.”

When you ditch the theories and look at actual numbers for Ohio, the idea of disenfranchisement gets ridiculous.

Ohio has 8.85 million voting age residents, according to the U.S. Census estimate of July 2011. A Columbus Dispatch article from the same month reported that 8.83 million voting age Ohioans had driver’s licenses or state ID.

Meanwhile, the state has 7.8 million registered voters.

RELATED ARTICLES

Technical Analysis of Ohio Voter Registration Files

Ohio: Comparison of Registered Voter Counts to Census Voting Age Population

Ohio Voter History: 77,000 voters registered in 2000 or before have no voting history

CONTACT: jon@ohiowatchdog.org