Turnout prompts concerns for Nov.-Election officials ask for more machines

[babylonsister]

Turnout prompts concerns for Nov.
Election officials ask for more machines

By Richard Wolf
USA TODAY


Record turnout in this year's presidential primaries has election officials worried about possible shortages of machines, ballots and poll workers in November.

In 17 of the 24 primaries held so far, turnouts were larger than any in the past 40 years, the result of competitive Democratic and Republican contests and earlier primaries. Paper ballots ran out from California to the District of Columbia, more poll workers were needed in Arizona and a voter registration database crashed in Connecticut.

"The biggest problem during the primary season has been too many voters," says Doug Chapin, director of electionline.org, which tracks voting issues. "Time and time again, the problem has been turnout being up higher than even the most optimistic projection."

Now officials from Virginia to Texas are warning that they will need more voting machines in the fall to avoid long lines. It's not clear whether county or state officials will pay for additional machines.

"This could be a tsunami coming," says Kimball Brace of Election Data Services.

more...

http://www.usatoday.com/printedition/news/20080229/1a_lede29_dom.art.htm

[Fly by night]

1. "The biggest problem during the primary season has been too many voters," WTF

Oh yes, too many voters -- that certainly must be the problem. (JFC).

The real problem is that, in addition to the many other problems associated with voting on DREs, they are an incredibly slow way to vote. The following is a comparison we did here in Tennessee of the efficiency of opscan voting vs. DREs. It is no wonder that Fortune magazine picked DREs as the worst new technology of 2003.
----

Two small examples of the inefficiency of DREs versus optical scan voting systems

1) Two identical groups of 21 voters are selected to vote either on an optical scan or a DRE. Given the size of the group, there would be no reason to have more than one piece of each type of equipment for them to vote on. (Besides, that keeps the cost comparison roughly equivalent.) For this test, each voter in each group is paired with a voter in the other group to spend exactly the same amount of time completing their ballot and having it counted. For the optical scan voters, that involves completing a paper ballot and having it read by the op scan. For the DRE voters, both completing the ballot and having it counted occurs on the same DRE.

For this example, each group of voters has a fast voter who completes the completing the ballot/counting process in three minutes. The voters then stair-step (in terms of time needed) so that each next pair of voters takes another 30 seconds to vote (so one voter in each group takes 3 1/2 minutes, one voter each takes 4 minutes, etc), all the way until the slowest pair of voters takes 13 minutes to complete their ballot and have it counted. (I am not calculating the time necessary to find the voters on the rolls and to sign the pollbooks. Let's just assume that these tasks take the same amount of time for both groups.) How much time does it take each group to complete all their ballots and have all of them counted?

Optical scan group (n= 21) -- 13 minutes DRE group (n= 21) -- 2 hours and 48 minutes

Why is the opscan group so fast and the DRE group so slow? Well, all opscan voters can start completing their ballots at the same time, and can have their ballots read when they are finshed, the only time they need to interface with the opscan machine. Since it takes so little time to read the opscan ballot, having 30 seconds between each voter means that there is no time needed to wait to have the opscan ballot read. While the slow voters are still filling out their ballots, the faster voters are having their completed ballots counted by the opscan. However, with the DRE, since the same piece of equipment is used to complete the ballot and to count the votes, only one person can complete the ballot and cast her/his vote on that piece of equipment at a time. So regardless of how fast or how slow individual voters are, they must wait until all the voters in front of them have finished before getting their chance to start filling in their ballots.

2) Once again, two identical groups of voters are selected, with 21 voters in each group. In this example, each group contains five voters who take 3 minutes to complete the ballot, five who take four minutes, five who take five minutes, five who take six minutes and a single voter who takes seven minutes. Once again, all other voting-related tasks are assumed to take the same amount of time. In this example, how long does it take each group to complete all their ballots and have them all counted?

Optical scan (n= 21) -- 7 minutes DRE (n= 21) -- 97 minutes

Since the opscan machines read ballots so quickly, one group of five voters (e.g., the 3 minute voters) could be fnished having their ballots read before the next group (4 minute voters) are finished completing their ballots and ready to have them read. But again, since the DRE only allows one voter at a time to complete her/his ballot and have it counted, everyone else has to sit on their hands until everyone in front of them in line to vote has done so.

Under these two examples, there are only two ways to speed up the voting with the DREs to complete the election in the same amount of time as with the optical scan:

a) buy more DREs (10-15 more) and hope that slow voters don't end up using the same DRE, or
b) have DRE voters get discouraged with the long wait and leave the polling place without voting.

So if the time it takes to complete an election were the only criterion for selecting voting equipment, opscan would win hands-down. Add to that the facts that opscans are much less expensive per unit of votes cast and they start and end with a paper ballot that is available for an audit or recount, the selection really should be a no-brainer. So what is our problem (really) here in Tennessee?
--------------------------

"Never argue with a man whose job depends on not being convinced." H.L. Mencken

"It is difficult to get a man to understand something when his salary depends upon his not understanding it." Upton Sinclair

[liam_laddie]

2. Opscans vs. punchcards

The BoE here - Hamilton County, OH - is now scanning absentee and early-voting ballots, but not "counting" until the polls close Mar. 4. Counting means taking the Hart Intercivic MBB...Mobile Ballot Box (PCMCIA-type 2 cards holding the scanned results...and SFAIK, executable programs) and inserting them into the tabulators. A scanner can scan 600 ballots per hour; our ballots are two-sided and will be longer in November. IIRC, punch cards can be read at a few thousand per hour, with the results fed into the central tabulators at a similar rate. Soooo, not only does it take a voter longer to fill in the bubbles on an op-scan ballot than to punch out the chads on a card, the process of recording the votes into a system-usable form takes quite a bit longer!
As I recall, the main issue with punch cards was the overvote. And of course hanging chads. I've always assumed undervotes were intentional and not a concern. Punch card blanks are about a half-cent or less each; printed and bar-coded op-scan ballots are at least 20¢ each. In some ways, we're going backwards.
Only two Ohio counties selected the Hart system, and it's the only electronic system with which I am quite familiar. We use one eScan per precinct and one eSlate (DRE) per polling location. For those counties stuck with all-DRE, and now having to provide a paper ballot to any voter who asks for one, my sincere condolences.
I concur with the Fly by night's conclusions.

[Fly by night]

4. With precinct-based opscan, votes are counted immediately so your time comparison w/ punch cards ...

... is not accurate as far as punch cards being faster. If you were comparing central-based opscan, I would agree. But w/ precinct-based opscan, votes are counted when they are scanned. So at the end of the voting day, there is nothing else needed to be done than obtain the vote count from the opscan. (Nothing, that is, except conduct the mandatory random manual audits of some number of precincts.)

However, I want to say I have no problem with punch cards and wish they had not been wrenched from a number of TN counties who did not want to stop using that low-tech voting method. I did find that punch-cards in TN had a pretty high undervote rate compared to other methods but that could probably be straightened out with more attention paid to the punching process.

In any case, down to and including counting colored marbles, whatever we do is immeasurably better than DREs.

[Kurovski]

3. "...too many voters."

Voter suppression just gets harder and harder all the time.