Robert M. McIver was registered to vote in Teller County in the last federal election, but he never showed up at the polls.
Ruth M. Favatella didn't cast a ballot, either.
That's because, according to the Teller County coroner and Social Security records, McIver and Favatella had both been dead for more than two years.
The two were among at least 15 deceased people who were listed as eligible to vote two years ago in the small mountain county west of Colorado Springs, according to an Independent examination of voter records. Some had been dead for only weeks or months, others for many years.
And that's not all.
In addition to voters who had died, Teller County's voter rolls in November 2002 were also bloated with hundreds -- probably thousands -- of names of other people who no longer lived at their registered addresses.
More than 30 voters also appeared to be registered more than once. As of last Friday, there were 16,843 voters registered in Teller County -- 12,961 of them are "active," meaning they recently voted. According to the state demographer, 17,148 people over 18 currently live in Teller County.
The county's apparent failure or inability to update its voter rolls has led to a dubiously high voter-registration rate: At the time of the 2002 election, the county claimed to have 16,135 registered voters -- amounting to an amazing 99 percent of its voting-age population at the time.
El Paso County, by comparison, reported a registration rate of 86 percent, equal to the state's average. (There are currently 355,761 voters in El Paso County -- 271,579 of them active. The state demographer estimates 406,826 people over age 18 currently live in this county.)
And yet, Teller County's case of voter bloat was far from the worst in the state.
In fact, 18 Colorado counties had registration rates greater than their voting-age populations. Topping the list was tiny San Juan County on the Western Slope, which claimed to have 618 registered voters even though only 457 people 18 years of age or older lived in the county at the time. That's a registration rate of a whopping 135 percent.
Such numbers don't just result in bizarre statistics. Having large numbers of nonexistent or "phantom" voters on the rolls can be costly for counties that use mail ballots because they spend money to send ballots to such voters.
Worse yet, inflated voter rolls can heighten the risk of abuse, some observers caution.
"If you have voting rolls that are inaccurate, it increases the potential for fraud and mischief," warned G. Terry Madonna, director of the Center for Politics and Public Affairs at Franklin & Marshall College in Lancaster, Pa. Madonna previously served on a legislative commission that examined "voter bloat" problems in Pennsylvania.
Election officials in Teller County and elsewhere in Colorado, however, downplay the risk of voter fraud.
There are no indications that any dead voters in Teller County have showed up at the polls, according to the county's top election official, Clerk and Recorder Patricia Crowson. Nor have the other "phantom" voters been casting ballots, Crowson said.
"Our rolls are probably inflated," Crowson conceded. However, "that doesn't bother us unless I start seeing them voting."
Crowson and other election officials also say the problem can't be fixed easily because state and federal election laws, intended to protect people's right to vote, make it difficult to purge the rolls.
"There's nothing we can do about it," Crowson said.
In a crucial presidential election year that already has people jittery over the potential of voter fraud or foul-up, bloated voter rolls are not unique to Teller County or other Colorado counties. Media reports from Pennsylvania, Florida and elsewhere have revealed registration rates in excess of 100 percent in many counties.
Madonna believes any county with a registration rate over 85 percent has large amounts of what some call "deadwood" on its rolls. "I just refuse to believe ... that any place has 100 percent of voters registered," he said.
Madonna blames the 1995 National Voter Registration Act -- popularly known as the Motor Voter law -- which made it possible for people to register to vote when getting their driver's licenses.
The law made registration so convenient that many people sign up even though they're unlikely to actually vote, Madonna says. And when moving to another county or state, few people bother to notify the clerk and recorder in their former county of residence that they've relocated or registered somewhere else -- so they remain on the rolls.
But Lloyd Leonard, a spokesman for the League of Women Voters, says Motor Voter isn't the problem. While making it easy to register, the law also requires counties to keep their rolls clean, Leonard notes.
"If some counties are over 100 percent [registration], they're clearly not following those rules," he said.
In El Paso County, election officials credit diligent purging efforts for their relatively low registration rate of 86 percent.
"We're very meticulous," said Susan Russo, the county's assistant election manager.
But other counties aren't doing the job as well.
Lisa Doran, a spokeswoman for the Colorado secretary of state's office, said she's aware that many counties in the state have inflated registration rates. "Some are better at working their lists than others," she said. "There are some county clerks that just flat don't do it. They don't purge."
Added Doran, "It's very difficult to purge. You have to really make an aggressive effort."
Though Teller County reported having more than 16,000 eligible voters during the November 2002 federal election, a regularly scheduled purge of the rolls later that month brought the official number below 15,000.
Even then, large numbers of "phantom" voters remained.
Of the 14,748 people still on the rolls, 6,181 were listed as "inactive" -- meaning they were eligible to vote but didn't exercise their right during the election that had just taken place. In tracking the whereabouts of 100 such "inactive" voters randomly selected from the list, the Independent verified that at least 21 percent -- more than one in five -- didn't live at their listed address at the time of the election.
That's almost 1,300 voters.
One of them, 33-year-old Stephen C. Barrington, was surprised when told that he was registered at an address on Highway 67 outside Woodland Park in Teller County.
"I've lived in El Paso County for the last seven years," Barrington said.
Moreover, although Barrington used to live in Teller County, he never lived on Highway 67.
"I have no idea why they'd have that [address]," Barrington said.
Kelly R. Allen, 40, who moved within Woodland Park in the summer of 2001, was also unaware that her address hadn't been updated.
"I've renewed my driver's license since then, and I'm pretty sure I've done my jury duty since then," she said. "So I guess I'm surprised."
The sample tracked by the Independent has a 10-percent margin of error, so the real number of incorrectly registered voters in November 2002 was likely somewhere between 11 percent and 31 percent -- in other words, between 680 and 1,916 people.
It could potentially be even greater, as the accuracy of another 14 percent of voters' addresses couldn't be ascertained.
In addition, at least 31 people appeared to be registered more than once. A number of female voters were registered twice under different last names, though their first names, dates of birth, addresses and phone numbers were identical.
For example, Heather Danell House, 27, of Florissant, was also registered under her maiden name, Heather Danell Sharp.
"I had no clue," House said when asked about it. She said she was surprised because she had the name on her driver's license changed three years ago. "I guess they didn't take the 'Heather Sharp' off" the voter rolls, she speculated.
In other cases, people were registered under different spellings. For instance, David D. Mortiz of Woodland Park was also listed separately as David Duane Moritz, with the same date of birth and address.
"I wasn't aware of that," Mortiz said when contacted.
On the roll for years
Crowson says a number of common factors contribute to the inflated voter rolls. "It's really no mystery," she said. "Every clerk in the nation probably has the same problem -- similar problems."
While it might sound simple to update voters' addresses when they move locally, or to delete them from the rolls when they die or leave the county, Crowson says it's not.
Federal law allows election officials to purge voters from the rolls if they receive official notice that a voter has moved away.
If a person moves to another county in the same state and registers to vote or renews his or her driver's license, the secretary of state's office will alert the election officials in the person's former county of residence. But if the voter simply doesn't re-register or get a new license, or if he or she moves out of state, election officials may never find out about it.
The election officials can also purge the names of voters who have died. In Colorado, each county's election office receives a monthly list of people who have died in the state. However, people who die out of state don't show up on the list.
And when election officials don't receive any notice that a person has died or moved, the "phantom" voter can remain on the rolls for years.
Officials can eventually purge such people from the list, but only after a series of steps prescribed by law. First, the person must fail to vote in a federal election. When that happens, the clerk and recorder's office sends out a notice in the mail asking if the person still wants to be registered. If the card comes back undelivered, the person must still miss two more federal elections before his or her name can be deleted.
In other words, a person could register in December 2002, move out of state shortly thereafter, and miss the federal elections in 2004, 2006 and 2008 before getting purged.
"It does seem like it's a long time, but certainly there's a purpose to it," said Marguerite Duncan, El Paso County's election manager. "The feds do not want us purging people and making it inconvenient to vote."
Voting from the grave
As for Teller County's many duplicate entries, Crowson says they most likely result from data-entry errors. For a while, the county election department wasn't able to catch and correct such errors due to a computer problem, which has since been fixed, she said.
However, Crowson says she isn't worried about duplicate entries because the secretary of state's office will notice if anyone tries to vote twice.
"It's a nonissue," she said.
And when it comes to voters who have died, Crowson said she and her deputy clerk routinely purge the rolls of people whose names appear on the monthly "death list" from the secretary of state's office or who have death certificates issued by the Teller County coroner. They also cancel the registrations of people who appear in local obituaries, she said.
However, before Crowson was elected clerk and recorder in 2002, the county election office didn't cross-check its list with the county coroner's death certificates, she said.
Whereas the Independent could verify the existence of only 15 dead voters on the rolls shortly after the November 2002 election, then-County Coroner Deb Smith complained a few months prior to the election that more than 30 dead people -- all of whom had been issued local death certificates -- were still registered to vote.
"There was no excuse for them not to remove them," Smith said.
As of September 2003, at least seven of those people were still registered to vote in Teller County. The Independent found people who had died as long ago as 1999, 1995 and 1993 registered to vote at that time as well.
Even if the county never received any notice that a voter died in 1995 or 1993, the voter should certainly have been purged by 2003 if the county were following prescribed procedures. By then, a person who died in 1993 would have missed five consecutive federal elections, and a person who died in 1995 would have missed four.
"I can't explain it," Crowson said when asked why such voters remained on the rolls.
Indeed, when it comes to cutting "deadwood" from Teller County's voter rolls, Crowson says her hands are tied. She maintains she's already doing everything she's allowed to do under federal law.
"Legally, we can't do more," she said.
But why, then, was Teller County's registration rate so much higher than the state average or El Paso County's?
"I don't know if I can answer that," Crowson said.
She speculated, however, that Teller County's highly mobile population might be a factor.
Many people work for a short time at the casinos in Cripple Creek and then leave, she pointed out. Other people move to Teller County in the summer but leave when they discover how tough the winters are, she said.
Others might be part-year residents. Some of these people register to vote, but they never notify the county when they leave.
Election officials in some other counties with highly inflated voter rolls cite the same factor.
"It's the transient population," said Dorothy Zanoni, the clerk and recorder for San Juan County, in explaining the county's 135-percent registration rate. "It really corrupts our voter registration."
Zanoni said many people move to Silverton -- the county's only town -- in the springtime to take seasonal jobs. When the tourist season ends in the fall, they can't find year-round work and end up leaving.
The county is small enough that Zanoni knows who most of the "phantom" voters are. She says she knows of at least 140 people on the rolls who don't live in the county.
"It does bother me, but there's nothing I can do about it," she said, citing federal law. "I cannot cancel them at my whim."
Many of the Colorado counties that had registration rates above 100 percent in 2002 were small. Jackson County in northern Colorado, with only 1,163 voting-age residents, ranked third in the state with a registration rate of 119 percent.
"I didn't realize [it] was that high," said Charlene Geer, the county's clerk and recorder.
Joan Roberts, the deputy clerk in Hinsdale County, said she knew that at least 100 people on her voter rolls no longer lived there. But she was surprised when told the county's registration rate -- the state's fourth-highest -- was 118 percent.
Located in southwestern Colorado, the county had only 625 voting-age residents in 2002, though it claimed to have 736 registered voters.
Does size matter?
Still, population size doesn't seem to be directly related to "deadwood."
Some of the lowest registration rates were also found in small counties. Crowley County, southeast of El Paso County, had the state's lowest rate at 57 percent, and neighboring Lincoln County was at 63 percent. Logan County, in northeast Colorado, had a rate of just 61 percent.
In the cases of Crowley, Lincoln and Logan counties, the rates are actually artificially low because all three have significant prison populations. For instance, prison inmates make up 25 percent of Crowley County's official voting-age population, but they aren't eligible to vote.
Other small counties that don't have large prisons also ranked low in registration, however.
And rates also varied greatly among large Front Range counties. Broomfield and Douglas counties had registration rates as high as 99.8 percent and 97 percent, respectively, while Denver's rate was only 81 percent and Adams County's was 72 percent.
The transient populations cited by election officials in Teller and San Juan counties also don't seem clearly tied to the high registration rates.
Eagle County -- home to Vail, with many seasonal jobs and part-year residents -- had a voter-registration rate of just 79 percent, well below the state average of 86 percent.
And though Summit County had the state's second-highest rate two years ago, at 120 percent, it wasn't caused by the hordes of people who descend on Breckenridge seasonally. The cause was actually a computer problem that stopped the county from purging its rolls for two years, said Vicky Stecklein, the county's election administrator.
The Miami scandal
Though Doran says "voter bloat" is a concern, she also says very few instances of voter fraud have been discovered in Colorado.
The Help America Vote Act, passed by Congress two years ago, will help cut some of the "deadwood" by creating a central voter registry for all of Colorado by the end of next year, she said. When it's up and running, the registry will immediately eliminate any dual registrations within the state, as well as the names of people who have died in Colorado. It will also be linked to the state's driver's license database.
However, the registry still won't make it any easier to purge the names of people who move out of state or die in another state.
Meanwhile, elsewhere around the country, some jurisdictions have taken extra steps to keep their voter rolls as clean as possible, the Miami Herald reported last year.
Miami-Dade County in Florida, stung by a voter-fraud scandal in 1997, regularly scans a register of people who have filed change-of-address forms with the U.S. Postal Service. And in Chicago, election workers were planning to walk door-to-door to get the most updated voter information.
Other simple checks can also be performed. The Independent identified numerous dead voters by comparing registration lists to the Social Security Death Index, which can be accessed on the Internet. The conclusion? Of all "inactive" voters who were born before 1930 and registered to vote in Teller County as of September 2003, nearly 16 percent appear to be dead.
Property records, available from the Teller County Assessor's Office, also show whether a voter has sold his or her house, indicating that the person might have moved.
Sometimes, just looking up a voter's current phone number and placing a quick call is all it takes.
In reality, the law allows "a variety of mechanisms" for dealing with bloated voter rolls, said the League of Women Voters' Leonard
"In well-run jurisdictions, it's not a problem," he said.
Contributor Wayne Young provided research support for this article.