05/17/2012 05:34 pm ET Updated May 18, 2012

Latino Voters Likely To Bear Brunt Of Florida's Voter Rolls Purge

Less than two months after a Florida effort to identify and purge ineligible voters from the state's rolls began, opponents say it seems likely to disqualify a disproportionate share of eligible Latino voters.

About 180,000 people -- a group roughly equal to the population of Tallahassee, Fla. -- are at risk of being purged from the state's voter rolls because they have been identified as possible noncitizens, the Miami Herald reported. Looking at a smaller sample of 2,600 suspect voters initially identified by the state, the newspaper found about 58 percent to be Latino. Hispanic voters constitute just 13 percent of the state's electorate, according to federal data.

Florida often grabs national attention because it's home to 11.3 million voters and wields 29 Electoral College votes. But Republican secretaries of state elsewhere, including Colorado and New Mexico, have also launched aggressive efforts to identify noncitizens on their voter rolls. Critics contend that many of these purges have relied on questionable methods and allow public officials to make inaccurate claims about the number of ineligible people who have actually obtained and used voting credentials.

The purges have significant legal and political resonance. Black and Latino voters have historically seen their right to vote restricted by law and by practice, making an amendment to the U.S. Constitution and, a century later, the 1965 Voting Rights Act necessary. Some states and counties, including five in Florida, remain subject to detailed federal oversight of their electoral systems.

Voters of color are also more likely to be registered Democrats. So voter roll purges that call into question the eligibility of large numbers of minority voters are generally thought to be helpful to Republican candidates.

"We are very concerned about a disparate impact on the Latino community," said Katherine Culliton-Gonzalez, director of voter protection projects for the Advancement Project, a nonprofit that works to protect voter rights. "We're particularly concerned about people who have naturalized. They have gone through all the trouble to learn English, pass an American history test, and invested a lot of money and time only, it seems, to be targeted."

The Florida Secretary of State's Office did not respond to a request for comment by deadline Thursday.

"Florida voters need to know only eligible citizens can cast a ballot and we're doing everything in our power to ensure that is the case," said Florida Secretary of State Ken Detzner in a statement issued this month.

In early May, Detzner announced that his office had identified at least 2,600 noncitizens who were registered to vote. The voters were picked out by cross-referencing the list of noncitizens who had been issued Florida driver's licenses with the state's voter rolls. The names were then sent to county-level election officials, who mailed letters to the suspect voters.

The letter indicates that the voter is illegally registered to vote and that the very act of casting a ballot would constitute a third-degree felony, said Culliton-Gonzalez. The letters also make clear that those who wish to appeal the state's finding have 30 days to provide evidence of citizenship or their names will be purged from voter rolls. The letters do not specify the documents required to prove one's citizenship status, said Culliton-Gonzalez.

There is no doubt that state election officials are required by federal law to make sure that voter rolls are clean and accurate, said Myrna Perez, senior counsel at the Brennan Center, a legal research and advocacy organization at New York University. But that work must be done with great care, after a detailed look at the records and well in advance of elections.

"It can very easily be done sloppily and done behind closed doors in an irresponsible and even illegal way," said Perez.

In 2000 and 2004, Florida removed thousands of allegedly ineligible voters from the rolls if the person's name and birth date were found to be an "80 percent match" with that of an ineligible voter, Perez said. So, for instance, if Michael Smith, born in December 1976, had a criminal record making him ineligible to vote, Mike Smith and Michelson Smith born in December 1976 were also removed.

The method is not only imprecise, but statistically likely to capture both eligible and ineligible voters, Perez said. Statisticians have a name for just how frequently people within a distinct group share the same birth date: the birthday problem. Names, like shoes and clothing, go in and out of style. So it is also relatively common for people born around the same time to have similar names. Between social trends and stats, what would seem to be rare -- finding people who share a similar name and a birthday -- turns out to be relatively common, said Perez.

As for the use of driver's license lists, Perez noted that since drivers may register to vote at the same state agencies where they renew or obtain a new license, some noncitizens may have been accidentally registered to vote. That number is likely small, she said, and the number of noncitizens who have cast a ballot even smaller. What is more likely is that during the six to 20 years that it can take for a person to become a U.S. citizen, the individual may obtain a driver's license. If driver's license data are not updated to indicate when a person becomes a citizen, then such a person may be falsely identified as an ineligible voter, said Perez.

Indeed, Florida newspapers have identified about a dozen people included in the initial list of 2,600 suspect voters who became citizens after obtaining a driver's license. Others have pointed out Americans born in Puerto Rico and Americans of Puerto Rican descent born in the mainland United States who were incorrectly identified as ineligible voters.

In Colorado, Secretary of State Scott Gessler has asked the Department of Homeland Security to identify which of the 4,000 noncitizens who have obtained state driver's licenses and are also listed on the state's voter rolls have since become Americans. Gessler insists that 2,000 of these individuals have cast ballots in Colorado. The federal government has refused to cooperate with his inquiry.

About 32,000 people have become U.S. citizens in Colorado during the period covered by Gessler's analysis, according to Perez.

The Brennan Center is preparing a public information request to learn more about the method Florida used to identify alleged noncitizens in the electorate, Perez said. The Advancement Project is considering legal action against the state, said Culliton-Gonzalez.