By Sam Levine and Alvin Chang, The Guardian | February 25, 2021
A well-connected conservative group in Wisconsin nearly succeeded in forcing the state to kick nearly 17,000 eligible voters off its rolls ahead of the 2020 election, new state data reveals.
The group, the Wisconsin Institute for Law and Liberty (Will), caused a national uproar in late 2019 when it successfully convinced a county judge to order the state to immediately remove more than 232,000 people Wisconsin suspected of moving homes from the state’s voter rolls. The state, relying on government records, had sent a postcard to all of those voters asking them to confirm their address, and Will sought to remove anyone who had not responded within 30 days.
Democrats on the commission refused to comply with the order, believing that they didn’t have the authority to immediately remove the voters and that the underlying data wasn’t reliable, and wanted to give voters until April 2021 to confirm their address before they removed them. Appeals courts intervened and blocked the removals; the case is currently pending before the Wisconsin supreme court. There were still more than 71,000 voters still on the list at the end of January who did not respond to the mailer (152,524 people on the list updated their registration at a new address).
But new data from the Wisconsin Elections Commission shows how disastrous such a purge could have been. And the dispute underscores the way fights over how states remove people from their voter rolls – often called purging – has become a critical part of protecting voting rights in America.
Across the country, Republicans and conservative groups have pushed for aggressive purging, saying it helps prevent fraud. Democrats and voting rights groups say the process can be done haphazardly, leaving eligible voters, particularly minority groups and students, at risk of being wrongly purged.
In Wisconsin, of the 232,579 people who were flagged for potential removal from the rolls in October 2019, 16,698 people – 7.2% of the list – wound up confirming they wanted to remain registered to vote at the same address. Nearly 11,000 of those people voted in the November election (Joe Biden defeated Donald Trump by just over 20,000 votes in the state).
“7.2% never moved. That is a huge error rate,” Mark Thomsen, a Democrat on the bipartisan Wisconsin elections commission, said during a meeting earlier this month.
“17,000 voters is a lot of voters,” said Ann Jacobs, another Democrat on the commission.
Richard Esenberg, Will’s president and general counsel, however, said the new data was actually evidence that Wisconsin’s process worked. “If the number is 7%, then I think it’s fair to say that the movers list was reliable for the purpose that it is being used for, ie, to ask voters to confirm their registrations,” he wrote in an email.
Wisconsin officials are still trying to understand exactly why so many voters were getting wrongly flagged. 2019 was just the second time the state used a report from the Electronic Registration Information Center (Eric), a multi-state consortium that uses records from the DMV, post office and other government sources to help election officials flag voters who may have moved. Thirty states and the District of Columbia belong to the consortium and the system is generally considered a reliable way of identifying voters who have moved.
Wisconsin, however, is exempt from a 1993 federal law that requires states to offer voter registration services at DMVs. That may be leading to issues in matching DMV and voter records in the state; voters who change a car registration to a different address but haven’t actually moved may be getting incorrectly flagged as movers, Meagan Wolfe, the administrator of the Wisconsin elections commission, said during a meeting earlier this month.
Figuring out whether or not the data Wisconsin is relying on is accurate is crucial because a state statute says that local election officials have to remove someone from the voter rolls if they have “reliable” information they have moved. Will says the data is reliable and so the voters must be removed. Democrats and voting rights groups say the data is not reliable enough to cancel registrations.
The Guardian contacted more than 200 voters who informed the state at some point over the last year or so they were still living at the same address after they were flagged as movers in 2019. Several voters said they had temporarily moved out of Wisconsin but continued to vote absentee in the state.
That’s what happened to Riley Freeman, a 23-year-old from Waunakee. In 2018, he asked the post office to begin forwarding his mail to his college address at Northwestern University, just outside Chicago, but continued to vote absentee in Wisconsin. He didn’t register to vote in Illinois, apply for a driver’s license or register his car there. The state flagged him as a mover, even though he was still eligible to vote in Wisconsin and wanted to do so. He voted by mail in 2020; had the 2019 purge gone through, he would have had to re-register to vote from college before he could vote by mail.
“I still kind of considered myself a Wisconsin resident who just happened to live in Illinois nine months of the year,” he said, calling the process “a little bit unfair”.
Carlos Martin Del Campo, a 20-year-old from New Holstein in north-east Wisconsin, was also among those flagged. Towards the end of 2019, he left Wisconsin to live temporarily in California with his father, but always intended to return to the state and vote there. By the time Wisconsin’s spring election in April came around, he was back in the state and voted in person at the polls after confirming to officials there that he had not moved.
“In my case I would see why I was flagged. But it just concerns me the potential for my vote not being cast was there,” he said.
It’s not clear why voters temporarily out of state may be getting flagged as movers.
“If National Change of Address has it listed as temporary, and not with other codes or other dmv data also indicating a move, then we take them off the list,” Reid Magney, a spokesman for the Wisconsin Elections Commission, said in an email. “But it’s also possible they checked the wrong box on the change of address form, or there was a data entry error by the post office. Or they had some other transaction at DMV regarding their vehicle, etc.”
Voters in non-white and low-income zip codes were all more likely to get flagged as movers subject to a potential purge, according to a Guardian analysis of state data. People living in those same areas were also likely to be wrongly flagged as movers.
Researchers found similar trends when they studied racial disparities the first time Wisconsin attempted to remove voters using Eric data. In 2018, 4% of the voters flagged as movers wound up casting ballots at the same address. Minority voters were twice as likely to do so than their white counterparts.
The study suggests that simply sending voters postcards to confirm their address is probably not the best way to identify who may have moved.
“It highlights the challenges in doing [voter roll] maintenance when people have unstable addresses,” said Marc Meredith, a professor at the University of Pennsylvania and one of the authors of the study. “Postcards by themselves aren’t gonna cut it.”
Esenberg and other defenders of the aggressive removals have argued that even if Wisconsin officials did erroneously remove some voters from the rolls, Wisconsin has same-day registration, which allows people to re-register when they show up at the polls. But several voters on the list also told the Guardian that they continued to vote absentee from an address abroad, casting ballots from places like the United Kingdom and Japan. Those voters are unlikely to have the option to re-register on election day, and it may be more difficult for them to get the necessary documents to prove their residency eligibility in Wisconsin.
“We’ve just gotten through an election cycle where the right in this country, conservative activists and legislators are practically apoplectic over garden variety election irregularities. But in the context of this issue, they seem to be very comfortable with a 7% error rate,” said Jon Sherman, an attorney at the Fair Elections Center who went to court to try to stop the removals last year.
“If a voting machine junked 7% of the ballots you fed it, I don’t think you would call that a reliable voting machine.”
# # #
# # #