Voting rights groups have urged voters to check their registrations with local officials.
In Michigan, some 33,000 voters were removed from the rolls in August, a figure that is far higher than the number of deaths in the state during the same period — about 7,100 — or the number of people who moved out of the state — about 4,400, according to data from the Postal Service.
In Colorado, some 37,000 people were removed from the rolls in the three weeks after July 21. During that time, about 5,100 people moved out of the state and about 2,400 died, according to postal data and death records.
In Louisiana, at least 18,000 people were dropped from the rolls in the five weeks after July 23. Over the same period, at least 1,600 people moved out of state and at least 3,300 died.
The secretaries of state in Michigan and Colorado did not respond to requests for comment. A spokesman for the Louisiana secretary of state said that about half of the numbers of the voters removed from the rolls were people who moved within the state or who died. The remaining 11,000 or so people seem to have been removed by local officials for other reasons that were not clear, the spokesman said.
The purge estimates were calculated using data from state election officials, who produce a snapshot every month or so of the voter rolls with details about each registered voter on record, making it possible to determine how many have been removed.
The Times’s methodology for calculating the purge estimates was reviewed by two voting experts, Kimball Brace, the director of Election Data Services, a Washington consulting firm that tracks voting trends, and R. Michael Alvarez, a political science professor at the California Institute of Technology.
By using the Social Security database so extensively, states are flagging extra registrations and creating extra work for local officials who are already struggling to process all the registration applications by Election Day.
“I simply don’t have the staff to keep up,” said Ann McFall, the supervisor of elections in Volusia County, Fla.
It takes 10 minutes to process a normal registration and up to a week to deal with a flagged one, said Ms. McFall, a Republican, adding that she was receiving 100 or so flagged registrations a week.
Usually, when state election officials check a registration and find that it does not match a database entry, they alert local election officials to contact the voter and request further proof of identification. If that is not possible, most states flag the voter file and require identification from the voter at the polling place.
In Florida, Iowa, Louisiana and South Dakota, the problem is more serious because voters are not added to the rolls until the states remove the flags.
Ms. McFall said she was angry to learn from the state recently that it was her responsibility to contact each flagged voter to clear up the discrepancies before Election Day. “This situation with voter registrations is going to land us in court,” she said.
In fact, it already has.
In Michigan and Florida, rights groups are suing state officials, accusing them of being too aggressive in purging voter rolls and of preventing people from registering.
In Georgia, the Justice Department is considering legal action against the state because officials in Cobb and Cherokee Counties sent letters to hundreds of voters stating that their voter registrations had been flagged and telling them they cannot vote until they clear up the discrepancy.
On Monday, the Ohio Republican Party filed a motion in federal court against the secretary of state to get the list of all names that have been flagged by the Social Security database since Jan. 1. The motion seeks to require that any voter who does not clear up a discrepancy be required to vote using a provisional ballot.
Republicans said in the motion that it is central to American democracy that nonqualified voters be forbidden from voting.
The Ohio secretary of state, Jennifer Brunner, a Democrat, said in court papers that she believes the Republicans are seeking grounds to challenge voters and get them removed from the rolls.
Considering that in the past year the state received nearly 290,000 nonmatches, such a plan could have significant impact at the polls. |
