Rhode Island news
R.I. election: The numbers don’t add up
01:00 AM EDT on Wednesday, May 2, 2007
How many people voted in the statewide elections last November?
A. 387,952
B. 390,340
C. 392,884
If you don’t know the answer, don’t feel bad; neither do state election officials.
The Board of Elections — which oversees the counting of votes — says 392,884 voters cast ballots in the Nov. 7 general election.
The secretary of state’s Elections Division — which oversees the state’s Central Voter Registration System and tracks who voted in which elections — counted 387,952 voters in the same election as of Jan. 16. In a new tally Monday, that number had risen to 390,340.
That leaves 4,932 ballots that were cast without a voter voting — or 2,544, depending on which number you use from the secretary of state.
The Providence Journal began examining the results of the November election this spring, after reporting last fall that the names of nearly 5,000 registered voters in Rhode Island appear on a federal list of dead people. The Journal sought to find out whether any of those dead people voted.
Digital Extra
The newspaper’s review found no evidence of systematic fraud by people casting ballots in the names of voters who had died. But it did find a voter tracking system susceptible to error that could throw into doubt the results of close elections.
As an example, East Providence Mayor Joseph Larisa Jr. lost his at-large council seat to Isadore Ramos by less than 20 votes. But, the Board of Elections tallied more than 100 ballots that did not have a corresponding voter in the secretary of state’s count, either in January or on Monday.
Robert Kando, executive director of the Board of Elections, attributed the discrepancies to errors in how voter tracking information was entered into the secretary of state’s computer system. “I don’t have the slightest inclination there was ballot stuffing.”
Kando said that comparing the number of votes cast to the number of voters who checked in at each precinct is not part of the process of declaring the results of an election. “We certify winners without doing that. That doesn’t mean we don’t validate our system by examining that.”
The process of tracking voters starts when they check in at the polls.
Before they receive a ballot, voters sign a poll book next to their name, address, date of birth and a unique bar code, like those used at checkout registers, that identifies the voters. After the election, canvassing clerks in each of Rhode Island’s 39 cities and towns use a scanner to register the bar code of each person who signed the poll book. The result of the scanning is a voter history file, which records the date and type of election and the precinct in which a voter votes, plus whether the vote was in person or by mail.
If a clerk skips bar codes — individually or even a whole page of the poll book at a time — voters will not get credit for voting. Similarly, if a clerk scans the barcode next to a blank signature line, someone who did not vote will get credit for voting.
November’s election was the first general election for which the poll book scanning system was used.
“This is really the first time out of the box,” Kando said. “I believe there are still some bugs in the system, but that’s to be expected.”
Even so, Kando said he was disappointed to hear that the discrepancy between the Board of Elections and secretary of state is more than 2,000 voters.
“Despite our best efforts, we periodically make a mistake,” he said. “There aren’t going to be any perfect elections.”
Kando and Janet L. Ruggiero, executive director of the secretary of state’s Elections Division, said some local clerks may still be entering voter information six months after the election. That accounts for why the secretary of state’s tally of voters rose between Jan. 16 and this week.
Kando conceded that, six months after the election, all of the voter information should have been entered.
The Journal’s search for dead voters revealed several other ways that errors wind up in the voter history file:
Voters inadvertently sign on the line for someone else with a similar name.
Voters sign on the line above or below the one on which they are supposed to sign.
In one instance in Bristol, a voter signed the poll book twice, in two different boxes. Town Clerk Louis P. Cirillo said he does not believe the voter was given two ballots, but conceded he has no way to confirm that. The problem in Bristol was that Angelo C. Stanzione III and Angelo C. Stanzione both were registered to vote from addresses on State Street. Cirillo said it appears that Angelo III mistakenly signed the box for the older Angelo, then signed his own box without his first signature being crossed out. Although he was registered to vote in the last election, the older Angelo did not go to the polls because he’s dead. He died in 1992.
In one precinct in Providence, the poll book — the only physical record of who voted — was lost before it could be scanned.
In that precinct, the state Board of Elections ruled that all 1,506 voters eligible to vote should be credited with voting, even though only 466 ballots were counted there. That means the secretary of state’s count on Monday of how many people voted is artificially high because more than 1,000 people who didn’t vote are recorded as having voted.
Kando said all the voters were given credit so that people who voted would not be declared inactive, a first step toward being removed from the voting list.
In its search for dead voters, The Journal compared the Death Master File from the U.S. Social Security Administration with the voter history file from the secretary of state’s office. The Journal initially found 24 matches between the two files. Then, a reporter went to the cities and towns involved and examined the poll books for precincts in question.
A match was defined as a person’s first name, last name and date of birth being identical on both lists. Plus, if a middle name was used on both lists, they had to be identical. If the middle name was missing on one or both lists, it would still be considered a match if the other parts matched.
In 10 of the 24 cases, the match was the result of dead voters being scanned, even though no one signed the poll book for that person.
Another five dead voters got credit for showing up at the polls because a legitimate voter signed in the wrong box.
Four of the 24 cases were false matches, where a live person with a common name — such as Mary Leonard — shared a birth date with someone who had died.
Another three initial matches were the result of people being erroneously included on the Social Security list.
That left a single instance, in Providence, where someone voted in place of a dead person.
Marcellino J. Cabral, who went by Popeye, died of cancer in 1989 at the age of 61. But someone voted in his name in November 2004 and November 2006, according to Providence records.
The city’s Board of Canvassers has sent a letter to Cabral’s son, Marcellino J. Cabral Jr., who goes by Butch, asking whether he can explain who has been voting in his father’s name. Laurence K. Flynn, the board’s executive secretary, said Butch Cabral has never been registered to vote in Providence. He said, depending on Cabral’s response, the matter may be turned over to the state police.
The Journal left a telephone message for Cabral with a woman who identified herself as his mother-in-law. He did not return the call.
Ruggiero said the secretary of state’s office will spearhead an effort to remove dead people from the voting rolls.
When The Journal first reported that nearly 5,000 names on the Social Security death list are registered to vote in Rhode Island, a spokesman for then-Secretary of State Matthew A. Brown said the secretary would obtain a copy of the list, and disseminate it to the cities and towns.
But Ruggiero said that has not happened. She cited the cost of the list; the federal government charges $1,725 for the file.
Ruggiero said her office will, instead, obtain a list of dead Rhode Islanders from the state Health Department as far back as computer records go, and update it on a monthly basis. Her office will make a similar effort to coordinate with the Department of Corrections to remove felons from the voting list.
More top stories
Most viewed yesterday
Patriots’ addition of O’Connell applies pressure on Cassel
Wide receivers, offensive linemen take their turn under the microscope
Cash discount gives gasoline retailers, customers a breath of relief
Most active surveys
Are you renting a summer cottage this year? Or not?
Storm report: What are you seeing?
What's your favorite breakfast/lunch place?
Are you able to watch highlights of the Super Bowl, or is it too painful?
Most e-mailed in the last 24 hours








