Tennessee Voting Anomalies - Not So Cute

Posted by Mary Mancini on October 21, 2006 under Uncategorized |

Unlike some people, I do not fear change. However, I do fear radical change in the hands of partisans without a verifiable safety net. This is why paper trails for electronic voting machines are a really good idea. I mean, I can’t think of any other reason why a politician or election official would not want to provide safeguards for an election, like paper ballots, unless they were rolling out the red carpet for election fraudsters.

Tennessee seems to have a particular problem this year in Shelby County where Diebold manufactured electronic voting machines, sans paper trails, were rolled out for use in the August primaries, experienced numerous problems, and were described by Jim March of Black Box Voting as “a hacking contest, open to whoever was most efficient and motivated to alter vote totals.”

Early voting in Shelby County for this current election seems to be heading in a similar direction.

WREG-TV in Memphis reported that two voters were given computer cards to plug into Diebold Election machines that were supposed to provide the correct ballots for each voters address. They didn’t. And when one went to get poll worker, “her voting computer screen went black, and then up popped the words ‘Your vote has been rejected.’” Poll workers, checking on the error, told her that their tallies showed that she had voted. The woman, who had used the machines on other occasions, says that she never finalized her vote.

After hearing her story, election official James Johnson said, “…we realized that the machine began to cancel out and that’s what it apparently started to do.” Whatever that means.

The woman was able to cast a provisional ballot but without a paper trail to compare with the electronic voting tally, how can we be sure her provisional ballot will match, and therefore count towards, the official vote count?

So why is Shelby County still using the Diebold machines after having so many problems in August? Well, all attempts to stop have been deflected. A lawsuit filed by Suhkara A. Yahweh, a local minister, and four other plaintiffs, that “contends the new machines are subject to tampering, that workers in some cases were not properly trained in the use of the machines and that the machines denied voters the right to cast secret ballots” was thrown out of court. U.S. Dist. Court Judge Jon McCalla ruled on Wednesday, October 18, the first day of early voting, that “granting an injunction, which would effectively put a halt to the election, would be an extraordinary step he did not believe was warranted.” The judge had canceled a hearing on the matter that was scheduled for the day before.

John Ryder, arguing for the Election Commission, said in court papers that the plaintiffs’ claims are “speculative and woefully deficient in that they are not supported by law of fact.”

But that’s not what Jim March of Black Box Voting found. March was asked by four Shelby County candidates (Shep Wilbun, Sondra Becton, Vernon Johnson and Otis Jackson) who brought a similar suit after the August primary to help secure public records from the Aug. 3, 2006 election. He found well-documented and widespread security violations. March obtained a copy of the Diebold GEMS database, along with the Windows event log and what he found was, in his words, “shocking”. “The sheer number of legal and security violations in the event log were horrifying, and it also showed that Shelby County — or someone — was accessing the file during the middle of a Temporary Restraining Order prohibiting this.”

Below is some of the evidence March found in “election official declarations and discovery documents obtained in litigation over a recent election using Diebold machines” as retrieved from the “Windows event log” of the GEMS server:

1) At 4:00pm the day after the election, somebody loaded a program called “JD Secure” into the GEMS server. “JD” stands for “JumpDrive” by the Lexar corporation. A “JumpDrive” is a small memory device, concealable in a closed fist, that allows rapid transfer of data. The fact that JD Secure was loaded means that data was not only moved around on a small, easily concealable memory device, but also that the contents of the memory device were password protected, further concealing what information was being transmitted. This violates system certification procedures and public/party observation laws (our right to transparent elections). It is difficult to conceive of a non-malicious purpose for concealment of data transfer (or manipulation) via password protection.

Four attempts to load the JD Secure software was documented across a period of about 15 minutes. So this loading of the memory device password application wasn’t accidental.

(One malicious use for the Jumpdrive with password encryption would be for a fraudster to hand-edit vote totals on another machine, spend hours to get it to total correctly, and then upload altered results back to GEMS the following day via the keychain memory device.)

2) On 8/7/06 somebody loaded a complete copy of Microsoft Office Professional 2000. This would include the MS-Access program long known as the simplest way to edit the contents of a Diebold central vote database (and banned for use in elections in virtually all jurisdictions nationwide).

3) On 8/22/06 for a period of over an hour — and during a time period for which a Temporary Restraining Order was in effect — somebody attempted to edit HTML data files. The only significant HTML files on a Diebold main server are vote total reports. The only reason any trace was left was because the HTML editing tools failed to load – they either broke somehow or weren’t loaded correctly. Either way, it suggests that somebody may have been hand-editing vote total reports, and that’s potentially very significant. For example: GEMS creates reports of vote tallies throughout election night, in either HTML or PDF formats. An editor for PDF data is included on the GEMS main server as shipped by Diebold, and HTML editors are easy to get. Those tally reports could be easily edited on election night, shifting the apparent vote totals, and then the main GEMS database could be hand-edited to match the false reports across most of the day after the election, to be illicitly uploaded via the password-protected Jumpdrive.

A Final Shelby County Fact (based on election official depositions):

4) At 6:30pm on election day, the GEMS main server was cross-connected to the main county computer network, which in turn has a cross-connection to the main Internet. They did this to allow uploading results from four regional elections offices across the county.

To be clear, it appears that results were modemed in to the regional stations from the precincts. The regional stations were reportedly connected to the county network.

It is impossible to overstate the seriousness of this security violation. This practice is flat banned in California and many other states, and may or may not be illegal in Tennessee…we’re checking on that. As a practical matter, this allowed anybody sitting at a county PC to get to and manipulate the central vote tabulator database.

With this final insult, it can be plainly stated that this wasn’t an election. It was a hacking contest, open to whoever was most efficient and motivated to alter vote totals.

And there’s more.

Now we have this. Early voting numbers from the first day of early voting (as reported by the Tennessean):

County — 2006 — 2002
Cheatham — 310 — 148
Davidson — 1,150 — 3,900
Dickson — 449 — 346
Hamilton — 2,669 — 1,452
Knox — 2,615 — 3,238
Maury — 632 — 610
Montgomery — 1,509 — 1,107
Robertson — 455 — 370
Rutherford — 1,721 — 939
Shelby — 8,148 — 584
Sumner — 571 — 555
Williamson — 1,056 — 694
Wilson — 761 — 652

Although the Tennessean labels the Shelby County numbers as “explosive,” and “14 times more than opening day 2002,” others are quick to explain the obvious discrepancy claiming that this is not necessarily an increase in total voters but simply an increase in voters voting early (due to 19 early voting locations). Although this could be a plausible explanation, because of Shelby County’s recent election history it’s way too early to discount other explanations, including foul play.

The lesson here - and from the 2004 election (remember the exit polls?) - is that not every election anomalie can be explained away and it’s up to us to continue to ask questions and demand evidence. Our job does not end on November 8th. Those of us concerned with election integrity need to fight against the “tin-foil hat brigade” label and fight for both paper ballot and Instant Runoff Voting legislation.