Onslow County N.C.failed to count 9,943 ballots on May 6, election night

[WillYourVoteBCounted]

Onslow County North Carolina failed to count 9,943 ballots on May 6, election night. That is 15% of all ballots cast.
Onslow uses 100% paper using ES&S M100 optical scanners and Automark ballot marking devices.
This had nothing to do with paper or lack of. This appears to be a failure to upload all results or to do so incorrectly. It would seem reconciling the election night materials would have caught this 15% discrepancy?? I don't have all of the answers.

Thousands of votes missed in Tuesday tallies

Same candidates win; commissioner vote rankings change. JD News May 9, 2008

Thousands of votes were omitted from election-night tallies, election officials have determined, but new unofficial results do not change which candidates won.
...
The Onslow County Board of Elections has corrected the number on the results Web site and is saying that no results were changed as to who won or lost the race.

The rankings among the candidates for Onslow County Board of Commissioners did change, however, as 9,943 more votes from the ballots were added. The change also considerably narrows the gap - from 82 to eight votes - between the top five vote-getters in the Republican race for those seats and the next runner up, who changed from Jack Bright to Jeff Jones - so a few provisional ballots could change that race's outcome.

N.C. Rep. Robert Grady, R-Onslow (District 15), said he filed a complaint with the North Carolina State Board of Elections on Thursday afternoon after he noticed a major discrepancy in the votes and could not reach Rose Whitehurst, director of the
Onslow County Board of Elections.

...Grady thought the votes were from the Jacksonville Commons' one-stop voting.
"People that voted before (May 2) with the one-stop voting didn't have their vote counted," he said.

Grady said he was told that the tabulator broke and the Onslow County Board of Elections then took the numbers from the computer and put them on a card.

"The votes from this card were not counted (Tuesday)," he said.

...more at the link

The vote totals before and after correction

Total previous results: 55,929

Total corrected results: 65,872

Difference 9,943

[RaleighNCDUer]

1. And that is why we have those voting machines.

They provide a paper trail, and if there is a discrepancy we can go to the original ballots and get it right.

[Wilms]

2. I didn't get that from the story...not that it's written very clearly. :(

Assuming the reporting is correct, it says that information on a memory card from a broken "tabulator" didn't get uploaded. Someone noticed...I'm guessing...differences between the number of voters and the vote totals. So they uploaded the memory cards again and got a result they feel is right.

They don't mention scanning or hand counting ballots.

Memory cards are used on Optical scanners and DREs, alike. A similar situation with a VVPAT-less DRE could have the same outcome.

[WillYourVoteBCounted]

3. I don't think the election director reconciled all materials

if she had, she would have realized there was a 9,000 vote discrepancy
compared to the poll books.

A candidate had to file a complaint and get the NC State Board of Elections
to investigate.

If one of the candidates had not complained, this director would not have
noticed.

The good thing is there are paper ballots there.

But the problem was the official not making sure everything added up.

[Wilms]

4. re: "The good thing is there are paper ballots there." Funny you should say.

Is anyone actually CHECKING those paper ballots??

I explained above that the story sounds as though there'd be no difference if the scenario included DREs. Is that not correct?

Too bad, rather, that memory cards were used in addition to the computerized vote counters.

Removing memory cards from the chain could improve the system security. (Think Hursti, UCONN, Brunner, and others.)


But, "the good thing is there are paper ballots there" that aren't being counted. :eyes:

[WillYourVoteBCounted]

5. answering the questions:

1. In Onslow, I do not think that ALL of the paper ballots were counted.
The discrepency was reported by a candidate.

2. I think that the scenario is the same as might happen with DRES in that either the software malfunctioned, or more likely - in this case - the election director malfunctioned and didn't upload the information correctly and exacerbated things by not performing her reconciliation procedures.

Lack of using proper reconciliation procedures could happen with hand counted paper ballots as well.

An exception from how this is like using DRES is this:
with DREs you depend on triple redundent memories that don't always match, and you don't have a real backup to refer to - the toilet paper trail has about a 7-9% failure rate in the ES&S iVotronics which are what NC certified.

One big difference with DREs verses Optical scan would be the option AFTER the fact - NC has and probably will again conduct recounts in optical scan counties and end up with different results, as has happened before.

Overall, I think that we have a case of someone not following the steps with the computers, but far worse, not reconciling the poll books, ballots cast, poll tapes and all election materials so that the discrepency would be found in the first place.

[Wilms]

6. Yes. That's largely my read on it.

From the story I didn't get that there were uncounted votes, though it wouldn't surprise.

And indeed you are correct to point out, As I should have, that the reported failures were of the type would effect HCPB elections as well as DRE and OpScan.

[WillYourVoteBCounted]

7. they're doing a recount today in Onslow

and the NC SBoE is going to audit that county BOE's office.

We will see if this is a lack of management,or what.

We also have problems with a count in Mecklenburg and Wake, I found out the
other day, but I am waiting for answer from NCSBOE as to which machines, software
etc were involved.

IN these counties, some of the early voting numbers were DOUBLED.