Tale Of The Tape:
With 80,000 New Votes Tallied, Election Night Accuracy Questioned
By Joseph Orovic
 |
| *includes affidavit and absentee ballots. |
Yards of receipt tape in folded piles or tightly-wound cylinders cling to returns of canvasses, large yellow sheets about the size of a standard desk calendar. Their numbers recount the tale of Nov. 2’s preliminary election tallies in pencil lead and ink. For one night, they fueled excitement, speculation and angst for many. And for 28 days, the numbers misled all.
When the City’s Board of Elections certified its poll results on Nov. 30, 195,005 electronic ballots were added to the state’s final tally, all of them seemingly lost in the Election Day shuffle of a mistake-prone counting system birthed by new voting machines.
Of that staggering figure, 31 percent, or about 80,000, belonged to our borough. The head-scratching disparity between the election night figures and certified results left many baffled. The preliminary vote-counting process remains wrought with so many traps, it could, for one very brief moment in one electoral district, leave an incumbent Assemblyman trailing behind write-ins like “Joy Behar” and “Peewee Herman.”
How could the numbers be so wrong?
Some ask a different question: Do we really need to know the results right away?
A New Process
In the now comical-seeming stone age of switches and levers, poll workers would open ballot machines at the end of the night, mark each candidates’ vote tally on a “Return of Canvas” sheet, then distribute them in three different directions. The system was wholly archaic, primitive in its use of the imposing 600-pound machines, and prone to error.
When the Help America Vote Act was signed into law by President George W. Bush in 2002, few in the state could imagine what would ensue eight years later. New voting machines, employing the latest fill-in-the-bubble technology, promised to make voters’ lives easier. Instead, a litany of complaints arose, with issues like lighting, privacy, readability and trust coming to the fore. It led Mayor Mike Bloomberg to reportedly call primary night “a royal screw-up” and left the Board of Elections scrambling to right any wrongs.
But the gap between preliminary and certified results highlights a wholly different issue – the plight of poll workers, volunteers and civil servants, learning a new technology while struggling to provide accurate numbers.
The new polling machines print out a reported 20-25 feet of receipt tape, providing duplicates of tallies. One set of results is posted for poll watchers. The other is cut manually with scissors according to election district.
The figures are then scrawled onto the Return of Canvas forms by the polling site’s inspectors, to be tallied and entered in the form’s “Total” column.
Two of three copies of the canvass form are then handed to a police officer, to be inputted into a database by a civilian employee of the NYPD at the precinct. Those inputted figures are disseminated as early returns to the masses.
The newly-found ballots were not enough to invalidated any concession or victory speeches. All of the borough’s Nov. 2 victors did, eventually, win.
With so many figures being transcribed, added and re-filed, the risk for human error, it appears, is massive. And in some cases, Total columns were left entirely blank.
“I’m not sure the public really appreciates the extent to which human error plays a role in election results,” said Assemblyman Rory Lancman (D-Hillcrest), who is proposing legislation that would let out-of-county college students work at polls.
The reasons for the errors are many, and agencies are quick to promise fixes.
Of Cops and Poll Workers
 |
| Return of canvass forms showcase just one of the many steps that could introduce human error into the preliminary vote tally. |
The 5 a.m. clock-in on Election Day left many poll workers frazzled. The day ahead of them was long. Questions had to be answered. Instructions were doled out ad nauseam. The road back home was at least 16 hours away.
When polls closed at 9 p.m., many a poll worker began the most tedious of the job.
Armed with the machines’ receipts, workers sat at tables and began transcribing figures, then adding them up with a calculator and cross checking the numbers. A coordinator would then inspect the sheets before handing them over to the police. At least that’s what was supposed to happen.
In some instances, faint receipt printouts rendered some numbers illegible.
“It was like the toner running out,” said Joseph Hennessy, who worked as a coordinator at PS 101’s polling site in Forest Hills Gardens. “Everybody complained about the printing.”
Also, workers relied on their own calculators – classic or cell phone – to add the numbers, if they even bothered.
At a recent stop by the Board of Elections’ office in Kew Gardens, the Tribune inspected three returns on canvass forms. Two of them did not have total vote counts added.
Prior to the Tribune’s findings, City Board of Elections spokeswoman Valerie Vazquez asserted that numerical snafus in the preliminary counts were not caused by flawed totals. She did not return phone calls for comment after the return of canvass forms were viewed.
The Board of Elections claims the Police Dept. did not properly input many totals at precincts on Nov. 2. In fact, 570 of the city’s electoral precincts initially reported zero votes in the gubernatorial race between Andrew Cuomo and Carl Paladino. The Board attributes that discrepancy to poor data entry on the part of the NYPD.
“Some have stated that maybe the police officer couldn’t differentiate what the number was so they put in a zero,” Vazquez said.
The NYPD did not respond to requests for comment.
And then there are the hours, grueling and long as they are. Some stayed beyond a full 18 in order to earn their $225. According to Hennessy, the time racked up and took its toll, an issue only worsened by employing a new system.
“The paperwork at the end of the night, that kind of complete switch around of what you had to do, was difficult,” he said.
The Gap Growth
The changed tallies solidified margins of victory in several key races and turned two casual wins into complete blowouts.
The victor for the 11th Senate district, Tony Avella, unseated incumbent Frank Padavan (R-Bellerose) by 3,083 votes, according to preliminary counts. That number jumped 38 percent to 5,003 votes when the certified numbers were released.
Avella said he did not worry that he might have actually lost when the new figures were released.
“Once you get over even 1,000 votes, it’s unlikely that newfound votes are going to change the election,” he said. “If it was under 1,000 votes, we might not have declared victory.”
Other incumbents faired much better. State Sen. Toby Stavisky (D-Flushing) gained more than 12,000 votes, while State Sen. Malcolm Smith raked in close to 19,000 more.
Does It Matter?
Despite the confusing figures, the preliminary numbers do not serve any legally binding purpose.
“Whatever the police put or anybody else puts out is irrelevant. It means nothing,” said Emanuel Gold, an attorney and former State Senator. “The official final results are what determine the election.”
Preliminary results are, in all truth, an opiate for candidates and the political-junkie masses.
Some questioned the necessity of preliminary results at all, saying the need for immediate figures and winners mirrors a distinctly American urge.
“We live in a society that wants instant gratification,” Avella said. “It’s not just election results.”
But these revamped figures should temper the Gospel of the early returns, according to Lancman.
“It just highlights the point that the election process is run by human beings,” he said.
The Board of Elections adamantly cautions against reliance on the early figures. The 200,000 additional ballots only served to prove their point.
“This is really the difference between official and not official,” Vazquez said.
Preventing A Repeat
Legislators like Lancman and Avella look to improve the election process through bills in their respective bodies.
“It’s just painfully obvious that the way we conduct polls should be carried out better,” Lancman said.
According to Vazquez, prevention of another preliminary gap will come with communication and exploration of flaws in the system. Poll coordinators are asked to keep journals which, along with a debriefing session, provide feedback to the Board of Elections.
The Board and NYPD will also meet and iron out any kinks, according to Vazquez.
“We want every voter to have every confidence at the end of the day,” she said.
Reach Reporter Joseph Orovic at jorovic@queenstribune.com or (718) 357-7400, Ext. 127.