There are two different punch card systems used in Florida, according to a story in the Washington Post today, written by reporter Dana Milbank. The differences between them helps to dispel some of the confusion that Republicans have been both exploiting and generating in order to obscure the numerous problems that have appeared disproportionately in Democratic strongholds.

The Votamatic system, used in urban Democrat strongholds counties like Palm Beach, Broward and Miami-Dade, produces hanging chads and other anomolies because it uses a paperclip-sized stylus to punch holes in a perforated ballot. The Datavote system uses a hole-punching machine with non-perforated ballots, which produces far fewer hanging chads (and presumably still less other anomolies, if any) according to the story in the Post. The Datavote systems predominates in smaller pro-Bush counties.

One punch card technology, the Votamatic system, involves punching holes in a perforated ballot using a stylus the size of a paper clip. This system produces the "hanging chad" that causes votes not to be counted by machines because the punched paper remains partially attached. This system is used in urban counties such as Miami-Dade, Broward and Palm Beach, which favor Gore by about 1,825,000 votes to 1,600,000. It allows more information on one ballot--an attraction in big counties with more items and candidates.

The other punch card technology, the Datavote system, uses a hole-punching machine on non-perforated ballots. This technology, though more expensive and more cumbersome, produces few hanging chads and, therefore, far fewer of the misread ballots that would be found in a hand count. This system is used mostly in small counties where Bush leads Gore about 90,000 to 60,000. It’s widely regarded as more reliable than the Votamatic system.

Other counties use either optical scanners or more primitive systems. Volusia County’s optical scanner system also malfunctioned.

To help put the Washington Post story in context, it should be noted that the county-level percentage of void presidentail ballots in Florida varied by a factor of 69 from a low of 0.18% in Leon County to a high of 12.4% in Franklin county. Palm Beach had a void rate of 6.43%, Broward 2.49%, and Miami-Dade 4.37.

Differences in void ballots between presidential and senatorial races cannot be explained in terms of the differences in voting systems. Nor do these differences in voting systems explain differences in double voting. But this story definitely does help to eliminate some of the confusion which has been so copiously generated over the past 10 days.

It is the major responsiblity of government to assure that every eligible voter is given the opportunity to vote. How astonishing it is that the State of Florida does not try to assure that their eqipment and voting system is free of blemishes. One would think they would rather see people of color and Jews not vote at all in as much as the State allows such poor systems to flourish in preciencts where the majority of voters belong to these groups. Of course the State would not do such a thing deliberately--would they?

Technology experts have been warning people about the Vote-O-Matic system since the mid-80s. This is not the current program making the rounds on geek sites, where you can vote over the web. This is an electromechanical setup dating from the real Jurassic Epoch of computing, the 1960s. It was sold to large jurisdictions by IBM, and represented a huge sale for them.

The equipment used to count the ballots resembles the punch card readers once used to enter code and data into the old mainframe computers of that era. These are the famous "IBM cards." They're more accurately called Hollerith cards. You don't see them much any more. They're kind of a symbol of old technology.

These readers weren't that precise to begin with, and several almost fatal problems were never completely resolved. The stylus machine does not cut a hole in the card like better equipment even at that prehistoric time. It just punches out a pre-perforated hole. This is asking for trouble. There's the dreaded "hanging chad." Even when this is gone, there's a little ridge where the card was perforated and this causes ballots to stick together, at which point both are considered void by the equipment.

Meanwhile, these clunky mechanical devices have had 35 years to get clunkier. Anyone who's tried to use an old punch card reader in a university (most have one around somewhere) know what I'm talking about. These were considered state of the art once. So was the Titanic.

A hand count is faced with the problem that there's no machine-human interface on the paper ballot itself. It's a friggin IBM card. This makes any hand count very subject to mistakes from fatigue, misreading, corruption, and human failing in general. Even so, it is still more accurate than a machine count with this system. Either way, the count kind of sucks (to use non-technical jargon).

The program code used to run the vote counting job is also old, and highly vulnerable to hacks. There's been nothing in the media about this. I think they either don't understand, or they do understand and realize that were the problems with this code ever revealed, that would be the temporary end of credibilty for democracy in many parts of this country.

Gore got -16000 votes on one machine count in FL. Computer geeks know that when numbers like this show up in the output, the code is in serious trouble. Either a count overran a variable and wrapped around, or a bit got flipped somewhere. One pro I know says he thinks that county was unsuccessfully hacked by Bush supporters, and that's all there is to it.

L.A. county uses the Vote-O-Matic too. You have to wonder.



Problems with the Vote-O-Matic system and computer voting in general:

Diamond, Larry; "Democracy Under Stress," Los Angeles Times Opinion Section, Sunday, November 19, 2000.



Jackson County, FL, abandons the Vote-O-Matic for optical sensing:

http://www.mailtribune.com/archive/99/aug99/81299n6.htm



San Francisco County, CA, does same. Also a discussion of problems with the Vote-O-Matic, with touchscreen voting, and why they too went with optical scanning:

http://www.ci.sf.ca.us/bdsupvrs/leganalyst/technology.htm

-- .- .-. .-.- --- -. ..