By Michael E. Miller
By Allie Conti
By Keegan Hamilton and Francisco Alvarado
By Jake Rossen
By Allie Conti
By Kyle Swenson
By Chris Joseph
By Michael E. Miller
One of the most intriguing mysteries of the whole Election 2000 debacle is this: How many Florida voters improperly lost their voting rights because of a statewide effort to scrub felons from voter rolls? This question was at the heart of a post-election lawsuit filed against the Department of State and others. The lead plaintiff, the NAACP, brought the class-action suit because more than half of those on the scrub list were black.
The good news is, all of those lawsuits are now settled. The private company contracted to perform the purge, Atlanta-based ChoicePoint (which in 2001 merged with the original contractor, West Palm Beach's Database Technologies, or DBT) has agreed to more closely scrutinize the names on the lists it sent out before November 2000 and identify those voters who should never have been removed in the first place. The supervisors of elections who wrongfully removed these voters from the rolls will then reinstate them.
The bad news? This unknown number of nonfelons (dozens? hundreds? thousands?) won't be back on the rolls in time to vote Tuesday. Some of them might already have been reinstated, and those who show up at the polls can cast a provisional ballot. But the original wrong -- the improper removal of their franchise -- has yet to be righted.
The NAACP brought the suit in January 2001 in federal court in Miami, alleging not only that the voter purge lists were flawed but that other violations of the federal Voting Rights Act had occurred in November 2000. The defendants were then-Florida Secretary of State Katherine Harris, DBT/ChoicePoint, and the supervisors of elections from seven counties, including Broward and Miami-Dade. Leon County settled first, in April 2002. Then came Broward in May, DBT/ChoicePoint in July, and Duval, Miami-Dade, and Volusia in August. The others settled the first week of September, on the eve of trial.
"The state was holding up the rear," rues Thomasina Williams, a Miami attorney who worked as co-counsel with the coalition of civil rights groups representing the plaintiffs in the NAACP case.
Was the Jeb Bush administration dragging its feet? "I'm reluctant to use the word delay," says attorney Elliott Mincberg of People for the American Way, one of the groups representing the plaintiffs. "Let's just say they litigated the case very aggressively, and there's no question that was a factor that stretched out the process."
Though the job won't be done until spring 2003, a second run through the original purge lists has arrived at a stunning conclusion. On July 29, ChoicePoint filtered its list of 94,282 potential or possible felons through more exacting criteria, including complete Social Security numbers. The number of possible felon matches: a mere 2,563.
This number is shockingly low. Was the list that DBT sent to the state actually 97 percent inaccurate? Probably not: The State of Florida does not routinely ask for Social Security numbers as part of the voter registration process. Still, Mincberg says his clients' expert was prepared to testify at trial that 70,000 of those 94,000 names were inaccurate -- roughly a 74 percent rate of "false positives."
Mincberg is generally pleased with the settlement terms, but he has reservations. "With respect to those who were improperly purged, that's not going to happen [in time for the 2002 election]," he says. He points out that the agreements, like all changes to election procedures in Florida, will have to be "pre-cleared" by the U.S. Justice Department before they are implemented. According to the Florida Attorney General's Office, the part of the settlement agreement that pertains to the felon lists was submitted September 26. As of press time, the Justice Department had not approved the agreements.
Florida is one of eight states that denies people convicted of felonies the right to vote and is the only state to include this denial in its constitution. This prohibition has had a clearly discriminatory effect: 27 percent of black men in Florida cannot vote.
A felon can ask the state to restore his or her voting rights. Not so long ago, the clemency process was almost automatic. In 1986, under Democratic Gov. Bob Graham, some 15,000 felons had their voting rights reinstated. In 2000, the last year for which figures are available, that number was 927. According to the Florida Parole Commission, there is a current backlog of some 26,500 applications for restoration.
The old system for maintaining accurate voter rolls went as follows: After checking with the Bureau of Vital Statistics and the Clerk of the Courts in his or her home county, each supervisor of elections would monthly strike the dead and the felons from the list, then send correspondence to counterparts in the 66 other counties noting the removals. This system was far from airtight, as the 1997 Miami mayoral election demonstrated. The most memorable lowlight of that mess was the fact that a dead man voted.
It was the state's attempt to rectify the Miami debacle that led to the most egregiously discriminatory aspect of the statewide debacle in 2000. The 1998 law established a central voter file and ordered the state to contract with "a private entity to compare information in the central voter file with available information in other computer databases, including, without limitation, databases containing reliable criminal records and records of deceased persons."
DBT was paid $4.2 million to make this comparison, then provide lists of felons to be removed. In 1999 and 2000 combined, DBT submitted to the state the list of possible felons. According to David Host, department of state communications director under both Katherine Harris and her successor, Jim Smith, the state whittled those lists down to some 71,600 names after removing duplicates and checking clemency records. The state then divided up the lists by county and sent them to the individual supervisors of elections.
The partisan context of this felon purge is particularly troubling. A Republican state legislature passed the law; the law was administered by a Republican secretary of state who was also the statewide campaign chair of her boss's brother's presidential campaign; the contract was awarded to a firm with strong ties to the Republican Party; and those voters who were targeted for removal would in all likelihood have voted overwhelmingly for Democratic candidates.
There were also some early glitches, such as the inclusion of 8,000 names of former residents of Texas who actually were guilty of misdemeanors, not felonies. But the most controversial aspect of the DBT list was how imprecise it was. Indeed, this lack of precision was exactly what the Department of State told DBT it wanted. George Bruder, a high-ranking employee of DBT/ChoicePoint, said in a deposition for the NAACP lawsuit that he had informed the state in 1999 that the list would result in "a significant number of false positives."
The state's response? "They wanted to make it broader and capture more names and let the supervisor go through the list and pull out those they could not verify," Bruder said.
Host concedes that the state had asked DBT, in effect, to cast a wider net in searching for possible felons. "The number of variations is extraordinary," Host states, citing both clerical errors and the frequent use of aliases by felons. "If you're going to do a thorough job... you have to go further than an exact match of names."
The breadth of this search had the effect of disproportionately targeting black voters for disenfranchisement. As first reported by London Observer columnist Greg Palast in a December 4, 2000, story for Salon.com, more than half of these putatively felonious voters were black. The list also included one supervisor of elections. Thus, by the time the state sent out the "felon" lists to the counties, the supervisors were highly skeptical.
"We'd heard from other counties that they'd found people on there who were upstanding citizens and shouldn't have been on there," Palm Beach County Supervisor of Elections Theresa LePore remembers. "We started to wonder, is this very accurate? We did some spot-checking, and we found more errors than actualities."
After these spot checks, LePore decided not to use the list at all, despite the probability that some ineligible felons would remain on the rolls. "I'm going to err on the side of the voter," she declares now. "In my opinion, let 'em vote."
Jane Carroll, the long-time supervisor of elections in Broward County who recently retired, also decided not to use the list, as did several of her counterparts statewide. "We realized early on that the information contained in that list was dubious at best," says Joe Cotter, who served as assistant supervisor of elections under Carroll and recently returned to administer the office under Miriam Oliphant. "We sent 4,800 certified letters out to people on the list; the responses we started getting back said that this was inaccurate. We began to suspect that this was not data that we wanted to use to remove people from the voter rolls."
In Tallahassee, Leon County Elections Supervisor Ion Sancho received a list of 694 names from the state. "Looking at the data they gave us, it was overbroad in its matching criteria." Sancho recalls. "I was shocked. This list was absolutely useless, worse than useless. I could see how it could lead to the illegal disenfranchisement of Florida voters. And it was clearly disproportionately negative to the African-American voters of the state of Florida." After assiduously checking the names on the list, Sancho and his staff removed only 34 felons.
But the supervisors in two large Florida counties took an opposite approach. In Miami-Dade, Supervisor of Elections David Leahy followed the same procedure that Carroll did in Broward: In both 1999 and 2000, his office sent out certified letters to every registered voter on his scrub list. According to documents filed in the NAACP case, the total number was 7,150. The 1,277 who responded got a hearing with the supervisor's office. Of these, 808 were determined to be felons; the other 469 either came up clean or had had their voting rights restored.
And what of the people who didn't respond? They were kicked off the rolls. "The law dictated that we do that," Leahy says. Of these, at least 3,357 were black. That's 62 percent. The population of Miami-Dade County is 20 percent black.
In Hillsborough County, supervisor Pam Iorio followed the same procedure. She recalls that she received "a little over 3,000 names" and ultimately removed about 2,300 people from her voter rolls. She says 54 percent of those on the list were black. "As I viewed the statute, we had to do something," Iorio says. "I cannot fault other supervisors for not using the list. In hindsight, that might have been the best move."
Responding to the flurry of complaints, lawsuits, and a U.S. Civil Rights Commission investigation that followed the 2000 election, the Florida legislature in 2001 passed the Florida Election Reform Act. Among other things, the statute provided for creation of a central voter database. That list is regularly checked for dead voters and duplications but is not being checked against the state database for felons -- and will not be until the Justice Department approves the process.
Another reform was the creation of the provisional ballot. In theory, this process would allow anyone scrubbed to vote on Election Day 2002. Iorio admits that a potential voter who hears that his name is not on the list might just walk away. "These are ex-felons," she says. "Maybe they're not eager for any confrontation with officials of the government."
Williams, the Miami attorney, says that provisional ballots are problematic, pointing out that if one is filed at the wrong precinct, it will not be counted. Even after all the litigation and publicity over this issue, Williams says, it is still quite likely that some of those who were purged in 2000 will once again be unable to vote in 2002.
The Department of State's Host says he believes the speculation about thousands of disenfranchised voters has been wildly overstated. "With respect to people who might have been erroneously removed, they've had two years to work that out with the supervisor," Host points out, noting that only two such individuals came forward during the NAACP litigation. "Our feeling is that any voter who has been affected by this has had a lot of notice to address this issue." Still, he adds, "One person who is not able to vote is too many."
Sancho doesn't believe it is the voters' responsibilty to fix the problem. "I think it's incumbent on elections officials to ensure that we don't compound the error of 2000 by disenfranchising legal voters again," he seethes. "We have a duty to allow every lawful voter to vote and to count that vote. We did not do that in 2000, and I, as an elections official, was embarrassed to be a part of the worst election, on a statewide basis, in the history of the United States. We have a moral responsibility to ensure, by whatever means necessary, that that does not occur again."