iOS app Android app More

`Rather Reports' on Vote-Count Fiascos

stumbleupon: `Rather Reports' on Vote-Count Fiascos   digg: US Works With Sudan Government Suspected Of Aiding Genocide   reddit: `Rather Reports' on Vote-Count Fiascos   del.icio.us: `Rather Reports' on Vote-Count Fiascos

FRAZIER MOORE | August 13, 2007 03:13 PM EST | AP

Compare other versions »

NEW YORK — With the 2008 election season heating up, familiar scapegoats continue to take the hit for past hang-ups at the polls. Those include bad graphic design (Florida's confusing "butterfly ballot" in 2000) and software glitches in certain voting machines.

But this week's edition of "Dan Rather Reports" explores other culprits: the very paper from which punch-card ballots were made, and glaring shortcuts in how certain touch-screen voting machines were produced.

"Our story is not that the election would have turned out differently in 2000 if certain things hadn't happened. No one can know that," Rather said Monday. But his eight-month investigation has "dug down vertically as deep as we were capable of doing" to probe the brewing problems _ including on-camera interviews with workers who had a front-row seat.

The hourlong news program premieres Tuesday at 8 p.m. EDT on cable's HDNet channel, with subsequent re-airings and streaming online video.

Rather's report begins with the current congressional bid by Democrat Christine Jennings, who lost her 2006 race by 369 votes in Florida's Sarasota County, where touch-screen machines showed 18,000 ballots with no candidate selected in that race.

How could that happen?

The broadcast hears from Gene Hinspeter, an electronic operations specialist in nearby Lee County, who speaks of a "calibration issue" with the touch-screen devices: on a misaligned display, choosing one candidate's name might actually trigger a vote for another candidate.

The touch-screen machines are hard to keep calibrated, says Hinspeter. He describes them as "unreliable."

While the touch-screens at issue were manufactured in the U.S., they are one of many components assembled in a factory in the Philippines.

Eddie Vibar, an electrical engineer who worked there between 1999 and 2002, describes the bare-bones performance testing ("They shook the machines"). He adds that conditions were oppressive at the factory, where the temperature sometimes rose above 90 degrees and only a few air conditioners were operative.

"It's hard to do repairs while you're also holding a fan or a piece of cardboard (to keep cool)," explains Vibar. He says he earned about $2.50 a day.

In a separate interview, Landen Tuggle, an American dispatched to overhaul factory operations, says that, despite his best efforts, 15,000 to 16,000 potentially defective voting machines were shipped to the U.S.

Rather's report also takes a look back at the fiasco that spurred the widespread changeover to touch-screen machines: the 2000 election, notably in Florida, where "hanging chads" and other irregularities caused havoc. In that state, more than 50,000 punch cards were discarded as invalid because voters appeared to have voted for more than one presidential candidate (or none).

Rather interviews seven former employees of the company that made punch cards used in Florida. They agree that after decades of maintaining high production standards, their company in 2000 began opting for cheap, even defective, paper.

"It's the flour for the bread," says one former worker. "I mean, if you don't have good paper, you won't make good ballots."

___

On the Net:

http://www.hd.net