Dawidziak: Voting problems surface again

Michael Dawidziak is a political consultant and pollster.

Before last week's Coliseum referendum completely recedes from consciousness, let's pause to consider "just one more thing" (as the late, great Peter Falk, as Lt. Columbo, used to say). Not on the vote itself, but rather on the reporting of the results.

The glacial pace of election-night returns left most media and political observers once again shaking their heads over the new voting machines. By 11:30 p.m. last Monday, only a little more than half of the precincts' vote totals were posted on the Nassau County Board of Elections website.

It would be difficult to overstate how excruciatingly slowly the results trickled in that evening. New technology almost always brings with it improvements in speed not previously thought possible. In this case, it doesn't seem possible that the results could have taken as long as they did to tabulate.

For comparison purposes, let's take a look at how election-night totals used to be counted on the old lever machines. At 9 p.m., when the polls closed, two Board of Elections workers would open up the back of each voting machine. They would painstakingly read off the totals of each column and row and write them down: "Row A, Column 1: 240 votes." And so it would go, until all the totals for every column and every row were recorded.

In any given election, there's a minimum of five parties, or rows, on the ballot; in presidential election years, there are many more. The number of contested offices, or columns, on the ballot varies from year to year, but there are usually at least 12. It can be over 20.

This means that workers had to read and record somewhere between 60 and 200 or more totals. Then they had to call the results in to the Board of Elections, where another worker typed them into a computer for tabulation and posting.

With our new system, when the polls close, the new machines automatically print out the totals for the workers. Last week's vote was a simple "yes" or "no" on a single question. With a total of 1,160 election districts in the county, this should have been a fast and easy process. A couple of high school kids with a computer and a spread sheet program could have tabulated the results in well under a half-hour.

Instead, after the polls close the police collect the drives from the machines and deliver them to the Board of Elections, where they are downloaded onto the main system -- a painfully slow process. It would be faster to call in numbers from the printouts, but when Suffolk did this last year, it led to initial incorrect results in the Tim Bishop-Randy Altschuler congressional race.

Nassau County Elections Commissioner William Biamonte, who fought implementation of the new system, has said, "This is dysfunctionality brought to an expensive art form."

In the old days, parties and campaigns used volunteers or "runners," who were present when the lever machines were opened. They would listen to the count and write down the results -- running them back to campaign headquarters. Campaign workers would then hand-tabulate the numbers. Usually, we'd know in an hour or so what the results were.

Ballot secrecy has been another repeated complaint about the new voting system. In the old machines, a voter was shielded on all four sides when casting a vote. With the new machines, a voter is shielded only partially, on three sides. And unless voters are careful, others can view the ballots as they are fed into the scanner. Last week, this was exacerbated, because there was only one question on the ballot. It was all too easy for others to see a voter's choice.

When the new machines were first used, New York City Mayor Mike Bloomberg said, "We've gone backwards." In an era when people expect to get their news instantly -- as it happens -- those in charge of running our elections need to figure out how to go forward.