Guest: Windham, NH auditor, Prof. Philip B. Stark of UC Berkeley...
In support of his lie that the 2020 election was stolen from him, Donald Trump has been telling anyone who will listen that "thousands and thousands" of ballots have recently been discovered from the 2020 race, including in New Hampshire, where the state's Republican Governor certified his loss to Joe Biden by some 55,000 votes. But something troubling was discovered during a post-election hand-count last November in a small New Hampshire town, revealing that the state's decades-old Diebold optical-scan computer tabulators undercounted some Republican votes and overcounted some for a Democrat on the ballot. Thanks to a real (non-clown show) forensic audit in Windham, NH, which is finally wrapping up, we now know why, as explained in detail on today's BradCast. [Audio link to full show is posted at the end of this article.]
On November 3rd, there were four seats up for grabs for State Representative in Windham, NH. There were eight candidates --- four Republicans and four Democrats --- running for those four seats. The top four vote-getters in the contest would win the seats. On Election Night, according to the computer tabulation, the four Republican candidates swept all four slots. One of the Democrats, Kristi St. Laurent, came in 5th by just 24 votes in the small town where some 10,000 votes were cast. So she asked for and received a hand recount about a week later.
The hand-marked paper ballots cast in Windham were tabulated on Election Night, as they are across most of the Granite State, by decades-old optical-scan devices originally manufactured by the now-defunct Diebold Election Systems, Inc., and long programmed and serviced by a company with a shady background named LHS Associates. The intellectual property of the systems themselves is now owned by the private vendor Dominion Voting Systems, which acquired many of Diebold's assets some years ago.
St. Laurent's requested hand-count, about a week after Election Day last November, discovered that the four Republicans in the State Rep's race each should have received another 300 votes or so, while St. Laurent actually was found to have received about 100 fewer votes than she was credited with in the Election Night computer tally.
So, what happened? How could the numbers have been so wrong? The NH state legislature subsequently adopted a measure to allow for a post-election forensic audit to answer those questions. It was a real audit. A professional one. Very much unlike, in almost every way, the long-running "audit" theater clown show still running in Maricopa County (Phoenix), Arizona. This one was run as a fully open and completely transparent public exercise over about three weeks in May, as organized and run by three voting system, cyber-security and post-election auditing experts tapped by the state.
We're joined on today's show by one of those three experts, PHILIP B. STARK, Professor of Statistics and Associate Dean of Mathematical and Physical Sciences at the University of California, Berkeley. He is also the inventor of the post-election Risk-Limiting Audit (RLA) protocol, regarded as the gold standard for post-election audits.
While the auditors are still wrapping up a few items before issuing their final report, the weeks of fully transparent recounting of ballots and physical audits of the scanners has yielded what Stark --- and the other auditors, including the legendary Finnish cybersecurity expert Harri Hursti and Mark Lindeman of the non-partisan Verified Voting --- regard as the answer to the mystery of what actually happened.
Via painstaking and extraordinarily well-document public experiments, they determined that folds in absentee ballot papers that went right through the oval next to St. Laurent's name resulted in the scanners incorrectly recording those folds as votes in many instances. That means, that if a voter had voted for all four Republicans on the ballot and the fold went through St. Laurent's "target", the systems may have seen that as an overvote in the State Rep's race and, therefore, wouldn't count any of the votes in that contest. Similarly, if the voter hadn't voted for four candidates and the fold went through the Democrats' oval, she might have picked up a vote that she didn't actually deserve.
"We're confident that we've identified the primary contributors to the mistabulation of the votes in the State Representative contest in Windham," Stark told me. "They are, namely, ballots that were folded through vote targets in contests, including in particular, Kristi St. Laurent's vote target in the State Representative contest."
"That, by itself, is not the only thing going on, because in the experiments that we subsequently did with the folded ballots, we found that there was a great deal of variability across machines and their ability to correctly interpret folds through vote targets, folds through ovals," he explains, noting that another problem discovered was dust on the optical eye of some of the very old scanners. "Maintenance does matter, because just using compressed gas to blow the dust out of the reader on the machine that was performing worst made it perform substantially better."
As Stark details, through experimentation, folding ballots in varying ways and by various means, the auditors were able to replicate the problem. "We have a pretty good understanding of how the ballots came to be folded the way they were folded, which was generally not along the score lines that the printer puts on them to guide where they should be folded. If they'd been folded in the right place, this wouldn't have been an issue."
Stark also explains why human eyes --- as in the hand recount last November --- are generally much better than computers at actually determining voter intent, even in normal cases, because scanners are simply not programmed to read certain types of marks made by voters as actual votes.
We also discuss the critiques and even threats from some Trump supporters that the auditors have endured from Trump supporters throughout the official audit, despite the auditors extraordinary, landmark efforts at transparency and public oversight, including answering hundreds of questions both at the audit site and via Twitter throughout the process.
"The complaints seem to be largely based either on a misunderstanding --- whether deliberately or not --- on the things we have said and a lack of understanding of the [Diebold] AccuVote OS [systems]; a lack of how New Hampshire runs its elections or understanding of New Hampshire election law; also, people somehow thinking Twitter is a definitive source of information about the audit, rather than the live [audio and video] feed and the New Hampshire Dept. of Justice's website, where we've been posting all of the evidence as we've accumulated it," Stark responds. "There are some people who are definitely accusing us of fraud, lying, of being on George Soros' payroll, of trying to cover something up."
"There are definitely people who think that Mark, Harri and I were hired to cover something up. I don't know what it is we are supposed to be covering up. The idea that we are not motivated to not find malware, if malware is there, is kind of laughable. It would be a career-making thing for us. We'd like nothing more than to find malware if malware is there! But we're not going to pretend it's there if it isn't."
Of course, Hursti himself has already made his career previously by hacking these very same AccuVote systems, back in 2005, as seen in the climactic finale of HBO's Emmy-nominated 2006 documentary Hacking Democracy and its chilling 2020 follow-up, Kill Chain: The Cyber War on America's Elections.
"There was no evidence of malware...And the effect that we've observed is just completely consistent --- every test that I've thought of to throw at it, every thing that we've done --- it really looks like it's the folds. Now, people are saying maybe somebody 'weaponized' the folds and knew exactly how to fold the ballots to take away votes from the Republicans, which is kind of the Wayne's World test: It could happen...and monkeys could fly out of my butt."
Stark also answers questions about how this audit compares to the very very different one ongoing in Maricopa. "The way that they're doing their hand-count is guaranteed to produce an inaccurate count," he charges. "You really need multiplicity and redundancy. You need to reconcile things as they go along. And putting ballots on a Lazy Susan and having them spin around in a hurry is not a way to get an accurate count."
Moreover --- and, of course, this is the only thing the disgraced former President actually cares about --- Stark speaks to whether or not this problem could have possibly affected other races on the ballot, including the Presidential contest.
Also on today's show: A few more thoughts, for now, on Sen. Joe Manchin (D-WV)'s obstruction of safeguarding democracy in the U.S. Senate, and Desi Doyen with our latest Green News Report with a whole lot of very disturbing news for the climate, and a longtime, genteel British environmentalist who is no longer pulling his punches on what he describes as the "crime" of climate change...
* * *While we post
CLICK TO LISTEN OR DOWNLOAD SHOW!...
The BradCast here every day, and you can hear it across all of our great affiliate stations and websites, to automagically get new episodes as soon as they're available sent right to your computer or personal device, subscribe for free at iTunes, Stitcher, TuneIn or our native RSS feed!* * *