Probe: Voting numbers don't add up   (CT)

BILL CUMMINGS      Connecticut Post    14 June 2008

In the midst of a heated court battle over last fall's Democratic mayoral primary, state Rep. Christopher Caruso's legal team asserted there were more votes than voters.

City officials and their lawyers scoffed at Caruso's contention, calling it untrue and irresponsible.

However, a Connecticut Post examination of election records from the Sept. 11, 2007, primary shows there were more votes than voters — 105 more.

Not enough to change the outcome — state Sen. Bill Finch defeated Caruso by 270 votes — but enough to raise some electoral eyebrows.

The Post found 105 more ballots were run through the optical scan voting machines than there were voters who checked into polling places. A total of 9,804 voters cast ballots for mayor.

At Park City Magnet School 249 people checked into vote and 280 ballots were run through the machine — 31 more than the number of voters. Park City is in the heart of Caruso's legislative district and is considered one of his strongest districts.

At Roosevelt School, 309 voters checked in to vote and 329 ballots were passed through the machine. "I've said from the beginning that this election was stolen," Caruso said.

"You only needed a few votes to swing that election one way or the other. When you put it all together, there was a conspiracy between the Democratic Town Committee, supporters of Finch and a willing registrar of voters," Caruso said.

City officials believe the primary was
Advertisement
fair, the results accurate and that Finch legitimately won.

"The vote is accurate. I'm really, really sure of that. Candidates were not affected by any of this," said Santa Ayala, the Democratic Registrar of Voters, who ran the election.

Finch said he's confident there were no significant problems during the election.

"As determined by the State Supreme Court, the minor irregularities in this primary were no different from any other election under either the new or old voting systems," Finch said.

"The Secretary of the State's office had staff on the ground on primary day in Bridgeport, and determined then and subsequently that there were no abnormalities," Finch said.

Adam Joseph, a spokesman for Secretary of the State Susan Bysiewicz, said it's not unusual to find a few more votes than voters in a given polling place because poll workers make mistakes, such as forgetting to cross out the name of a voter given a ballot.

But Joseph said the discrepancy found by the newspaper was too large to be considered "usual" and easily attributed to simple mistakes. He declined to comment further, citing the ongoing state elections commission investigation.

Almost immediately after he lost the primary, Caruso challenged the election in Bridgeport Superior Court, and later before the state Supreme Court. Although both courts rejected Caruso's allegations of a flawed election, neither court examined his contention that there were more votes than voters. Caruso failed to allege that in his complaint filed in Bridgeport Superior Court.

The issue may be reviewed by the State Elections Enforcement Commission, which launched its own investigation after the state Supreme Court ruled against Caruso in February. The elections commission cannot overturn the results of the primary, but it can issue fines and other punishments.

Nancy Nicolescu, a spokeswoman for the elections commission, confirmed the commission's investigation is continuing but declined to say what specifically it is examining. More votes than voters is something investigators "would consider," but she declined to confirm whether they are. She urged the newspaper to forward its findings to the commission.

Ayala conceded there were problems during the election, the first her office had run, and the first time the city used optical scan machines. She said future elections will run smoother.

Exactly what the 105 ballots represent, and how the discrepancy should be interpreted, is difficult to answer. The newspaper found no evidence that ballots had been illegally passed through voting machines; nor was there evidence that didn't happen.

It's possible the 105 ballots represent mistakes by poll workers who failed to check off voters as they entered the polls. If so, that's a violation of election law and raises questions about who cast the ballots. Voting law is designed to ensure that only registered voters vote, and only one ballot is given to each voter.

Caruso said he believes someone passed illegal ballots through the machines in a planned effort to increase Finch's vote totals. "With the lack of safeguards in that election, it made it ripe for fraud," he said.

Finch urged city residents to contact the state elections commission if they have concerns about voting irregularities.

"My office has taken steps to ensure that all Bridgeport polling places are opened on time and that backup poll-opening measures are in place for all future elections and that all election procedures are followed to the greatest extent possible," Finch said.

After Caruso lost the party primary, he appealed to Bridgeport Superior Court to overturn the election. The trial court found discrepancies and problems in the way the election was run, but concluded those issues were not sufficient to overturn the results.

The state Supreme Court ruled the same.

It pointed out that the city's registrar of voters violated election laws by failing to appoint moderators 20 days prior to the election, failing to provide the names of poll workers for public inspection, inadequately staffing polls and assigning poll workers to multiple positions.

It was in trial court that Caruso's lawyers claimed there were more votes than voters, and attempted to introduce evidence to back up the contention. Bridgeport Superior Court Judge John Blawie refused to consider the issue, ruling the allegation was outside the scope of Caruso's original complaint.

He also never specifically asked for a recount, which doomed his later effort to raise questions about discrepancies in the number of people who checked in at polls and the number of ballots run through the voting machines.

Instead, Caruso's team detailed dozens of alleged violations of election law as evidence that the results could not be trusted, including allegations of polls that didn't open on time, allowing only one person to supervise a polling place, improper signatures on voting forms, inadequate staffing of the polls and insufficient training of poll workers.

But none of those allegations stirred Caruso's supporters as much as the idea that there were more votes than voters.

The Post found that in 12 polling places, the number of ballots run through the voting machines exceeded the number of voters who checked in to vote.

In the remaining 11 polling places, the Post found more people had checked in than voted. Although that sounds strange, election officials say it's typical some voters will check in and, for whatever reason, decide not to vote. The Post arrived at 105 more votes than voters by counting the number of voters who checked in at each of the city's 23 polling places and comparing that number to the "machine count" of ballots successfully passed through the machine.

Ayala, the Democratic registrar of voters, said she is concerned "Someone forgot to do something" but "A lot of the forms were new to people."

She suggested jammed ballots counted by the machine caused the discrepancy. "It will show as a ballot that went through. The ballot can stop going through and have to extracted," Ayala said.

She said that while a flawed ballot would not be included in the number of votes recorded by the voting machine, it would show up in the machine's count of total ballots run through. Joseph, the secretary of state's spokesman, disagreed. When a ballot jams, a message flashes on the machine's screen, Joseph said. If the ballot is counted, the message indicates so, and the ballot, along with its votes, is included on the machine's tape, which tallies all votes and provides a total of the number of ballots successfully passed through the machine.

The jammed ballot is then extracted from the machine and placed in a container underneath, where ballots which successfully pass through end up for storage, Joseph said.

Those ballots are a backup to the machine tally of votes.

If the ballot jams and was not read, the message indicates it was rejected, Joseph said. The poll worker then removes the ballot and gives the voter a new one to fill out. The first ballot is placed in a separate container for "spoiled ballots," ballots from which votes were not recorded.

A "spoiled" ballot would not be included in the machine count of total ballots processed.

Besides problems in the number of votes versus voters, the newspaper found dozens of mistakes and errors in paperwork poll workers filled out. Those workers are required to add up the number of people who checked-in, along with the number of restorations and transfers. R&Ts are voters whose voting rights were restored at the polls or were transferred to a new polling place.

Only rarely were those numbers added up correctly. In fact, poll workers more often than not used the machine count of ballots as the check-in count. As a result, the paperwork, which is supposed to compare the machine count to a hand tally by poll workers, incorrectly reported a match between the number of people who voted and ballots that passed through the machine.

Many poll workers used pencils to cross off voters names as they checked in, a violation of election law. Pencil marks can be erased, which is why pens are mandated.

At Harding High School, forms filled out by poll workers claimed there had been 369 restoration and transfer requests at the school. When the poll workers added that number to the number of names allegedly crossed out on the voting list, they listed 769 voters as having passed through the polling place.

The problem is the machine count showed only 375 ballots had been run through the voting machine. Despite the fact that their own record keeping indicated 375 votes were cast by 769 voters, the various forms were signed, certified and sent to the registrar of voters office.

The Post determined that 377 voters had actually checked in at Harding, compared to 375 ballots cast. That meant two people checked in but didn't vote.

Ayala acknowledged the Harding numbers were "impossible" and represented a mistake. She said the entire election was short poll workers, and her office had only a month to prepare because a new state law changed how elections were run. That law took effect just over 30 days before the Democratic primary.

Town clerks for years ran elections in Connecticut, but the new law placed the responsibility on the registrar of voters, who previously had registered voters and maintained voting lists.

"It's a system we are perfecting slowly. We had 30 days to switch everything. I think considering the time we had we did a great job," Ayala said.

"The purpose of these laws is to have checks and balances. And when there are 105 more votes than voters, you are open to fraud," Caruso said.