Inside Job

by Steven Rosenfeld

  That rebellion was a remarkable development in the usually compliant world of election officials. But Florida is not just any state when it comes to purges. It got a black eye after the 2000 election for a catalog of dysfunction. One of its worst episodes involved Florida Secretary of State (and George W. Bush campaign cochair) Katherine Harris. She hired a contractor, Database Technologies, to winnow the state’s voter rolls.169 The National Voter Registration Act lays out what counties are supposed to do, which, needless to say, Florida didn’t do. The NVRA says no registered voter is to be removed unless they have not voted in four years—two full federal cycles—and only after postcards were mailed to them advising them they were being placed on inactive lists and could be purged if they do not resume voting.170 The 1993 law also says infrequent voting is not a sufficient reason to be removed from official lists, creating an ambiguity that’s come under GOP attack—but more on that later.

  What Harris did was piggyback on another racist legacy, Florida’s lifetime voting ban for felons. While denying the vote to felons has constitutional precedents, it has been twisted in a handful of states to disenfranchise an outsized percent of nonwhites—especially non-violent drug offenders in the war on drugs. In 2000, more than eight hundred thousand Floridians, about 7.5 percent of its voting-age population, had permanently lost their voting rights this way. (The figure in 2016 is double that: 1.6 million citizens, including 21 percent of voting age blacks.171) Harris hired Database Technologies to create a list of “possible” and “probable felons” who were on Florida’s rolls. It presented her office with 82,389 names, which was 1 percent of its electorate. In turn, she instructed county election officials to remove the voters who Database Technologies identified.

  “The tallies were wildly inaccurate,” wrote Michael Waldman in his book, The Fight to Vote. “The search used primitive methods, such as identifying voters who share a common name and birth date with someone who had a criminal record.” Thousands of legitimate voters were wrongly purged in the race where the Supreme Court stopped a recount and George W. Bush “won” by 537 votes.172 In his book, Give Us The Ballot, Ari Berman said that “no one could ever determine precisely how many voters who were incorrectly labeled felons were turned away from the polls.” He notes a legal settlement between the NAACP and Florida re-ran Harris’s purge list with “stricter criteria” and found twelve thousand errors.173 More than poorly designed ballots—where hundreds of Palm Beach County Democrats mistakenly voted for Pat Buchanan—the purge was a bigger factor in tilting the odds to Bush via a hidden administrative process.

  Fast forward to 2012 when Detzner revived Harris’s script with a twist. He swapped her fear of felons with his fear of noncitizens. Two years before, Detzner, whom Republican Gov. Rick Scott appointed, claimed that there were 180,000 possible noncitizen voters based on state driver’s license records. After pushback by voting rights advocates that number shrank to 2,600, and then to 200, and finally 85 people were removed from Florida’s voter rolls. In 2014, Detzner tried again. This time, he wanted to use a federal Department of Homeland Security database to identify noncitizens, even though the agency said it was incomplete. While Detzner and DHS argued in court, his county election supervisors refused to accept purge lists from Florida’s state election director. Their routine list maintenance process worked, they said. They didn’t need a solution for a nonexistent problem.174

  You would think in the relatively small world of election administration that such brazen moves would be known and widely rejected. But across the country, local election officials rarely are that bold—or some agree with the GOP’s goals. In mid-2016, Kim Strach, the North Carolina Board of Elections executive director, also appointed by Republicans, used a PowerPoint demonstration to augment her testimony before her GOP-run legislature. It said 35,750 voters were “registered in North Carolina and another state and voted in both [states for the] 2012 general election,” Greg Palast noted in August 2016’s Rolling Stone.175

  Where did her number originate? In part, the Interstate Crosscheck program run by Kobach. Under media scrutiny, Strach backtracked. “We were not suggesting that 35,750 voters had committed any type of fraud,” she told Palast. “There were 35,750 people who voted in North Carolina whose first and last names and dates of birth matched persons who voted in the same election in another state.” Strach was using porous data and knew it. “Despite hiring an ex-FBI agent to lead the hunt, the state has charged exactly zero double voters from the Crosscheck list,” the Rolling Stone article said.

  Palast then quoted Mark Swedlund, a “database expert whose clients include eBay and American Express,” who was stunned after looking at Crosscheck’s Georgia and Virginia lists. “God forbid your name is Garcia, of which there are 858,000 in the US, and your first name is Joseph or Jose. You’re probably suspected of voting in 27 states,” he said. Their point is partisan purges based on flimsy data are another large-scale means to take out eligible Democrats when the number of purged voters can exceed margins of victory.

  What’s emerged since Palast’s report is that North Carolina, through Crosscheck, had identified 455,891 suspicious voters as of February 2016. In Ohio, it identified 386,092 suspicious registrations. In Michigan, it was 406,268. In Arizona, it was 240,277. In the twenty-five participating states analyzed in 2016, more than 5.9 million registrants were flagged as suspicious by this Republican-led fishing expedition that pretends to protect the integrity of the vote as a supposedly effective administrative bookkeeping tool. These figures are from a March 2017 brief filed at the Supreme Court by Republican former Justice Department officials, including those who led the Bush administration’s voter fraud witch hunts. They urged the Court to hear a case over Ohio Republicans’ mass purge of hundreds of thousands of infrequent voters between 2012 and 2016.176

  In June 2017, the Court took the case, which will be heard in their fall 2017 term. In A. Phillip Randolph Institute v. Husted, Ohio’s Republican Secretary of State Jon Husted is appealing a September 2016 federal appeals court ruling that found the purges he ordered violated the NVRA. Husted created a “so-called ‘Supplemental Process,’” the ruling said, to identify all voters who had not voted in two years and put them on a track for removal.177 The Appeals Court said Ohio should not have purged infrequent voters—as many people only turn out in presidential elections every four years. A dissenting opinion said the NVRA was “ambiguous,” because, on one hand it laid out a process to remove voters who did not show up for four years—two federal cycles, but on the other hand it also said voters could not be removed “by reason of the person’s failure to vote.”

  This ambiguity was seized and exploited by Husted—just as Republicans in other states seized upon the lack of a definitive federal citizenship database to demand paper proof of citizenship. Under Husted, Ohio has a notorious record for purging voters. As Ari Berman noted in The Nation after the Supreme Court said it would hear the NVRA case, “From 2011 to 2016, Ohio purged 2 million voters from the rolls—1.2 million for infrequent voting—more than any other state.”178 The purges have unabashedly favored Republicans. In June 2016, Reuters reported that twice as many voters were purged in “Democratic-leaning neighborhoods” compared to “Republican-leaning” neighborhoods in the counties where Ohio’s three biggest cities are located—Cleveland, Cincinnati, and Columbus.179

  “The disparity is especially stark in Hamilton County, where affluent Republican suburbs ring Cincinnati, which has one of the highest child-poverty rates in the country,” Reuters reported. “In the heavily African-American neighborhoods near downtown, more than 10 percent of registered voters have been removed due to inactivity since 2012. In suburban Indian Hill, only 4 percent have been purged due to inactivity.”

  This disparity is another facet of what twenty-first-century voter suppression looks like. Democrats are targeted and peeled away, a few percentage points at a time. After the Sixth Circuit Court of Appeals ruled against Husted in September 2016, 7
,500 of the purged Ohio voters were restored. (That’s because the NVRA prohibits removals within ninety days of a federal election.180) But that victory only restored a fraction of the otherwise eligible voters who were removed. It underscores how the GOP benefits by throwing the most destructive first punches in a voting rights war of attrition.

  When thinking about Ohio’s purges, Crosscheck’s use in North Carolina, Georgia’s use of imprecise voter verification databases, or Florida’s bid to seize any government file to find noncitizens, one must not only ask why are Republicans doing everything within reach to obstruct voting. One also must ask why are voter file procedures and systems so prone to partisan abuses? Why are they so outdated and unsophisticated compared to private-sector data mining, which tracks people’s identity, address, and contact information, and lets no online search go unrewarded without deluging the searcher with banner ads?

  Part of the answer is politicians don’t want to change a system that they mastered—and that serves them. That is also true with the federal campaign finance system. As a result, elections often are kept as a backwater in state and federal budgets. They are underfunded and have aging voting machines that are hardly ever replaced. That takes on added significance when we look at the computer systems used—some of which were targeted by Russians in the run-up to 2016’s Election Day.

  But mostly, Republicans don’t want high turnout elections—and that starts with registration. In 2010, I briefly worked with the Pew Center on the States as part of its Voter Registration Modernization Design Working Group. It was clear that states could use better data to verify registrations—such as changes of address from credit rating agencies and the post office.181 It was also clear states could find and register all eligible voters if there was the political will—which Oregon later did. Pew devised a plan for a far better data-mining operation than Crosscheck that top state election officials could convince legislatures to use. It was a way to make rolls more accurate, efficiently managed, and cost-effective. But it wouldn’t automatically register voters, even though it identified all of the eligible but unregistered voters in participating states.

  Red state election directors said they could only take eligible voters up to the finish line, but those people would have to cross it on their own. They talked about citizens taking the final affirmative step. That’s as far as sympathetic Republicans went. Under Pew’s blueprint, the participating states would identify all their eligible voters and contact them to urge them to register. It also used better data to maintain voter rolls under the NVRA’s timetable. The system it created, called ERIC—Electronic Registration Information System—was used by twenty states in 2016.182 Some of these states are also in Crosscheck’s network, but do not use Crosscheck’s data due to its shortcomings. ERIC was a reason why 200 million people were registered to vote in 2016. (Notably, no ERIC states are represented on Trump’s election integrity commission led by Kobach.) But registration alone is not the key to winning. REDMAP showed it’s not just about candidates and registrations, but getting a reliable base to the polls in sufficient numbers. Clinton’s presidential campaign showed you could not alienate a sizable slice of your party’s base and get all of them to turn out.

  Predictably, as the 2016 election traversed from summer into fall and Clinton led in the polls, Trump began making repeated claims that the election would be stolen by voter fraud. “More than 1.8 million deceased individuals, right now, are listed as voters,” he said, citing Pew’s work.183 Of course, that Pew study never made claims about the dead voting; it was intended to show dead weight that needed to be removed by modernizing how voter rolls are maintained—part of making a case for ERIC. But, as the right-wing blogosphere is wont to do, that statistic was distorted and exaggerated. By January 2017, Trump’s Press Secretary Sean Spicer said, “There’s one report that came out of Pew in 2008 that showed 14 percent of people who voted were non-citizens.” That Pew issue brief, also created to promote the need for ERIC, did not mention noncitizens either.184

  But facts don’t matter when it comes to the voter fraud meme: double voters, dead voters, noncitizen voters, whatever. As a former Florida Republican Party chairman, Jim Greer, told the Palm Beach Post in 2012, voter ID laws and cutbacks in early voting were “done for one reason and one reason only,” to suppress the Democrats. Greer remarked Florida’s Republican political consultants “never came in to see me and tell me we had a fraud issue. It’s all a marketing ploy.”185

  America is nothing if not a marketing society. It’s unclear how much Trump’s revival of the voter fraud meme motivated his voters, apart from inflaming anti-Clinton sentiments among swing voters or preaching to the GOP’s converted. A poll in mid-September 2016 from the Washington Post–ABC News found that nearly half of registered voters across the country believed voter fraud occurs “somewhat” or “very” often. As The New York Times editorialized about that finding, “Another 26 percent of American voters said that fraud ‘rarely’ occurs, but even that characterization is off the mark. Just 1 percent of respondents gave the answer that comes closest to reality: Never.’”186

  By October, Trump wasn’t just tweeting, “Stop Crooked Hillary From Rigging This Election!” He was trying to recruit “Trump Election Observers” to police the polls. As in past presidential years, Trump was sending out a call for GOP loyalists to go to precincts in blue strongholds and challenge the credentials of anyone who didn’t appear to them to be a legitimate voter. There’s a long history of Jim Crow-era incidents of this ilk. But beyond the GOP putting up billboards in recent cycles in minority neighborhoods in battleground states187 threatening arrest should someone not present the proper voter ID—another lie—this is more bark than bite. Still, there were a few overt voter vigilantes in 2016.

  VOTING VIGILANTES

  MOST POLICE-THE-POLLS THREATS REMAIN JUST THAT—THREATS to discourage voting. But North Carolina saw a notable vigilante effort as the 2016 Election Day approached. A longtime right-wing activist, Jay DeLancy, who runs a group called Voter Integrity Project of North Carolina, targeted thousands of voters to be purged from three counties with large black populations. What he did is called caging—a tactic the Republican National Committee was caught doing in the 1990s, and signed a federal consent decree agreeing to stop. In 2013, DeLancy told a local weekly newspaper, “I can’t be a Republican and do what I’m doing.” But, of course, he is a Republican—and an activist in a state where party and race overlap.188 DeLancy said TrueTheVote, a Tea Party-based group that started in Houston, Texas, and sees itself as a polling place posse, had inspired him. TrueTheVote subsequently failed to go national in 2012.

  Delancy’s group specialized in a tactic that’s an electoral mugging. They sent postcards to addresses of registered voters in three key counties—people who had not voted recently. The postcards said, “do not forward” and thousands were therefore returned. They took those cards to county election offices as “evidence” to challenge those voters’ registrations. The NAACP heard about this and sued. One week before the 2016 election, a federal judge ordered the counties to restore 6,700 voters.189

  US District Judge Loretta Biggs ordered the State Board of Elections to restore voters who had been purged within ninety days of the fall election.190 Needless to say, county election boards knew the NVRA’s procedures—it’s been law since 1993. They should have ignored DeLancy’s vigilantism. However, the voter suppressor, in remarks that echoed Kobach’s ongoing proof-of-citizenship gambit, hewed to a state’s rights line. “The state did nothing wrong with following the law. We did nothing wrong in following the law,” DeLancy told the Associated Press. The wire service wrote, “he hopes the federal voting rules that Biggs cited in her ruling are ultimately challenged.” DeLancy had plenty of company among North Carolina’s voter suppressors. At the GOP-run State Board of Elections, staff attorney Josh Lawson said it would comply with the court order but might “appeal facets of the ruling, potentially after the election.”
<
br />   Sloppy partisan purges, using bad voter data in accusatory ways, taking preemptive steps to deprive others of voting rights—these tactics are taken for granted by fervent Republicans in North Carolina and elsewhere. They persist despite forceful court rulings in the past two years, starting with rejecting race-based gerrymanders in North Carolina, Texas, Alabama, and Virginia. Federal Courts have overturned other anti-voter laws in states like Wisconsin, Pennsylvania, and Kansas. But North Carolina stood apart in 2016—even surpassing Texas, where courts have struck down laws as racially discriminatory and Republicans keep trying to reinstate them.191

  In July 2016, a federal appeals court threw out five North Carolina anti-voter laws that its opinion said “target African-Americans with almost surgical precision.” These included a more restrictive voter ID requirement, reduced early-voting options, and fewer registration opportunities, from eliminating same-day registration to blocking teenagers from registering before they turn eighteen. All were adopted after the Supreme Court gutted the Voting Rights Act in 2013.192 Less than a year later in spring 2017, the Supreme Court refused to hear the state’s appeal—saying it would reserve the right to return to those issues another time, but not in a state with such messy politics. Later that term, the Court ruled in two cases that North Carolina’s redistricting was illegally racist—for two House seats and twenty-eight state legislative seats.193

  What unfolded in North Carolina as the 2016 election crested was as brazen as any of the Republican actions to seize or hold power this decade. It showed that as the pendulum swung back toward voting rights, the GOP had no reluctance to impede nonwhites and Democrats from voting, to disqualify their votes, and to overturn results. North Carolina’s story is even more remarkable because before REDMAP targeted it in 2010, it had the most progressive and inclusive voting laws in the South. It was politically purple, but the Republicans viciously rolled back the clock to turn its complexion red.