-- By Brad Friedman and John Gideon
New Jersey's Department of State, which at the beginning of this month took over all state-wide election duties from the state Attorney General's office, may be celebrating its new duties by sipping down a few glasses of Sequoia Voting Systems-flavored kool-aid in regards to the continuing saga of the Sequoia AVC Advantage touch-screen voting systems which failed during, and after, the state's Super Tuesday primary.
In a press release issued late last week --- which Sequoia was all too happy to selectively feature on its website --- the NJ Dept. of State made a couple of curious, and indeed misleading, announcements. Among them:
That, despite the finding of a NJ judge last Tuesday declaring that failed voting machines in some six different counties were to be subpoenaed by plaintiffs and subjected to mandatory independent review in order to determine the reason why at least sixty machines reported voter totals on their end-of-day paper tapes which disagreed with the internal numbers reported by the machines.
After the judge's order last week, Sequoia quickly moved to try and quash the subpeonas [PDF, 82 pages], despite the fact that they are not actually a party to the long-running court battle between the state of NJ and citizen plaintiffs suing them in hopes of getting rid of the touch-screen systems altogether (or, in lieu of that, have paper-trail printers added to them).
As Pennsylvania prepares to use the very same flawed Sequoia AVC Advantage voting machines in next week's crucial primary, we've tried to sort out what is --- and isn't --- going on in this New Jersey mess. We've also tried to determine who is reading the disputed paper tapes correctly, since a Princeton scientist seems to read them one way, while the SoS office and two of the NJ counties (which had originally read them as the Princeton Prof did) have suddenly decided to read them another way.
What we can tell you indisputably, however, is that if it's debatable as to what those paper tapes actually say, then those voting systems are in violation of the federal law which mandates "a permanent paper record with a manual audit capacity" that "shall be available as an official record." If there's a question about what the "paper record" from each machine actually says, than it certainly can't be used as an "official record" for auditing purposes.
To sort out the mess, we've touched base with the plaintiff's attorney in the case, as well as with the NJ SoS. Sequoia's spokesperson, VP and part-owner (for the moment) Michelle Shafer, on the other hand, doesn't seem to want to talk to us anymore for some reason...
Shafer recently stopped responding to several related queries after a BRAD BLOG exposé revealed some embarrassing details about her company's announced plans to commission their own "third-party" "independent review" of NJ's machines, by a firm named "Kwaidan Consulting." As we discovered after their announcement, Kwaidan was little more than a guy named Mike Gibbons of Sugarland, TX, who had a somewhat embarrassing personal history (hint: the history involves both "well endowed blonde nymphomaniacs" and "Jesus Christ").
In the wake of our report, exposing babe-magnet Mike and his attempts to cover his tracks, we subsequently learned that Gibbons was, in actuality, already a Sequoia contractor collecting a regular and rather hefty payday from Sequoia. He was also discovered to be a long-time friend and one-time employer of Sequoia VP and part owner Edwin Smith. The same Edwin Smith who had, days earlier, sent an email to a couple of Princeton computer science professors threatening them with legal actions should they perform a real independent review of the state's voting machines as requested, unanimously, by an association of county election officials in NJ.
The BRAD BLOG has since learned that, following our report, Gibbons was quietly removed from the project.
Ordered but Unsigned...
Without Shafer to help, we were able to get some clarity after a call to the NJ Dept. of State to ask about the disparity between newspaper reports such as last week's Star-Ledger and a very thorough report in Monday's Daily Princetonian (which picks up on on our amusing coverage of Gibbons, by the way) that the machines had been subpoenaed, and their press release indicating they hadn't.
How could they make the claim that "there is no Court Order requiring any county to release voting machines to the plaintiffs" in light of the fact that Sequoia Voting Systems themselves had already filed an effort to quash the subpoenas? After checking with superiors, spokesperson Shawn Crisafulli wrote back to explain it was all a matter of wording:
Any one can serve a subpoena, which can be challenged by way of a motion to quash. Here, the Sequoia Voting machine company has filed a motion to quash and the court will have to rule on the motion.
The SoS would appear to be splitting hairs here, with what would appear to be a carefully parsed followup to their original, unqualified --- and somewhat misleading --- direct statement declaring "there is no Court Order requiring any county to release voting machines to the plaintiffs."
Superior Court Judge Linda Feinberg has, contrary to the statement above, ruled that machines are to be turned over to the plaintiffs. Subpoenas have, as reported in the media, been issued by the plaintiffs. But the SoS is correct, that the court's ruling has yet to be signed, as the quash motion remains to be decided.
For further explanation of the murky mess, we contacted the plaintiffs' attorney in the long-running NJ case where last week, Feinberg reportedly ordered the independent review of the Sequoia touch-screen systems. In response to our query, Penny Venetis, Clinical Professor of Law, Clinical Scholar, and Associate Director of the Constitutional Litigation Clinic at Rutgers University, emailed us the following explanation on the current legal status of the Sequoia Voting Machine inspection:
The State said that the counties have the voting machines. In court, the State also said that it was not aware of any confidentiality agreements that prevented the machines from being turned over.
To get the machines, we sent subpoenas to the counties, because the counties are not part of the lawsuit. This was just a formality, given the judge's order that we can have the machines.
Instead of moving to quash the subpoenas, the counties did 2 things: 1) say that the machines were under the AG's control, so they won't turn them over to us. If this is the case, then we automatically get the machines through discovery! 2) Bergen County tipped off Sequoia and sent Sequoia a copy of the subpoena. Sequoia moved to quash. We are responding to the motion.
The Tale of the (Unreadable) Tape...
Also in question in the NJ Dept. of State press release are statements that they support the findings of Bergen and Gloucester Counties with regards to still more errors reported by Felten on the paper tapes he was provided by those two counties and others as a result of FOIA requests.
In addition to the original mismatch of voter totals found in six counties, Felten reported on his blog the new tapes "show more serious discrepancies than we saw before." Including scans of the tapes in his blog item, Felten showed several instances where there were more votes than voters reported.
But NJ disputed Felten's findings in their press release, claiming:
It was noted in the blog that one machine from Gloucester County recorded three votes for Ron Paul. However, county records only show two votes, which is consistent with the overall vote count.
In the case of the Bergen County machine, which was alleged to have registered 49 votes for Hillary Clinton, in fact there were 48 votes recorded. Unfortunately, it looks as though Professor Felten rushed to judgment and released a statement that is inaccurate. A misreading of the tape appears to be the source of Mr. Felten's error.
In other words, NJ is claiming that the audit printer tape failed to print the full character "8" on the tape, thus leading to Professor Felton's "misreading" the number as a "9." Here's a close-up of the tape posted on Felten's blog. Compare the "9" in Clinton's vote total with the 9s and 8s on the right side of the tape (click it to enlarge):
The tapes were sent to Felten by the two counties after he'd requested those which showed discrepancies. As he now writes in response to the state announcement, and the apparently new position of the counties involved, alleging that the tapes don't show descrepencies after all: "In short, the Secretary of State talked to the counties, and then the counties changed their minds about how to read the tapes."
Last week, based in part on the many reported failures of the NJ Sequoia Advantage touch-screen machines, VotersUnite.org's Ellen Theisen guest blogged on vote counting error rates far in violation of what is allowed by HAVA.
If Felten and the counties (and likely many of you who also) originally "misread" the above as "49" instead of "48," it would underscore yet another way in which the Sequoia AVC Advantage machine appears to be in violation of the Help America Vote Act (HAVA), regardless of whether the digit is, in fact, an 8 or a 9.
Yesterday, Theisen pointed this out, sending us the following section 301(a)(2) of HAVA [emphasis added]:
···(2) Audit capacity.
······(A) In general.--The voting system shall produce a record with an audit capacity for such system.
······(B) Manual audit capacity.--
·········(i) The voting system shall produce a permanent paper record with a manual audit capacity for such system.
·········(ii) The voting system shall provide the voter with an opportunity to change the ballot or correct any error before the permanent paper record is produced.
·········(iii) The paper record produced under subparagraph (A) shall be available as an official record for any recount conducted with respect to any election in which the system is used.
"If the printing is unreadable, or even disputable," Theisen wrote us yesterday, "how can the paper provide 'a manual audit capacity'? How can it serve as an 'official record'? It appears that, based on other (non-official?) records, NJ officials have now decided the 'official record' produced different totals than they thought when they first read it."
Did we mention these systems are to be used in two PA counties next week? Is your head exploding yet? Ours did long ago.
And who will bother to enforce the above-described HAVA violation? Well, that would be the DoJ or the U.S. Election Assistance Commission (EAC), of course. Feel free to begin holding your breath now.
In the meantime, NJ has $20 million in the pipeline to purchase new add-on "voter-verifiable paper audit trail" printers from Sequoia, wherein every voter can enjoy such confusion for themselves. That is, if the EAC ever federally certifies those printers for use. That, instead of throwing out these crappy machines entirely.
For the record, Sequoia's Shafer is quoted in the Princetonian, shamelessly and opportunistically echoing the NJ statement to take a bitter shot at Felten, claiming he is "exactly the type of reviewer that Sequoia wishes to avoid one that already has pre-determined the outcome of his investigation before it truly begins."
No doubt, she'd have preferred Sequoia's chosen --- then disgraced and dismissed --- reviewer, the Jim Beam-swigging Mike Gibbons.
By the time you read this article, however, there may not even exist a Sequoia Voting Systems any more, and Shafer may soon be out of a job entirely. Today is "Match Day" for the much-beleaguered company, when they must come up with $2 million in order to keep themselves from being subsumed in a hostile takeover by their much smaller competitor, Hart InterCivic, as The BRAD BLOG reported exclusively last week.
Is all of this any way to run a representative democracy?