[UPDATED at bottom of article following GOP primary run-off.]
When Stephen Colbert's older sister, Elizabeth Colbert Busch (D), faces her likely opponent, former Gov. Mark Sanford (R), in the May 7th U.S. House special election in South Carolina, it will all take place, once again, on the state's oft-failed, easily-hacked, always 100% unverifiable touch-screen voting systems. [NOTE: Unfortunately, despite today's date, this post is not an April Fools joke.]
The candidate announced as the "winner" of that election, based on those unverifiable voting systems, will either be the one who received the most votes from SC voters in the 1st Congressional District --- or the one who didn't. There will never be any way for anybody to know for sure one way or the other.
For that matter, after Sanford faces off with Curtis Bostic in Tuesday's Republican primary run-off election for the nomination, neither candidate will be able to know for certain whether they've actually won or lost that election either. This is South Carolina "democracy" in 2013 (and in at least one third of the country elsewhere.)
To make matters still worse, the state's Legislative Audit Council (LAC) has just released a new official report on those failed voting systems. Unfortunately, the report includes a enormous flaw which, when I pointed it out to Perry K. Simpson, Director of the LAC on Friday, his response was: "I see your point"...
Like a bad rerun
You may remember those SC voting systems as the ones used to 100% unverifiably record and announce that Alvin Greene --- an unemployed, unheard of candidate, who didn't campaign and had no website or even cell phone --- had "defeated" the four-term state legislator and circuit court judge Vic Rawl to win the 2010 Democratic primary nomination for U.S. Senate. The virtually illiterate Greene would go on to run against, and reportedly lose to, the incumbent Republican Senator Jim DeMint. It is DeMint's subsequent early retirement at the end of 2012 which led to the open House seat in SC's 1st district, after its former occupant, Rep. Tim Scott (R), was appointed by Gov. Nikki Haley (R) to DeMint's old U.S. Senate seat.
Shamefully, South Carolina still uses those 100% unverifiable ES&S iVotronic voting systems for every election run across the Palmetto State. And now, a new report on the state's voting machines issued by the South Carolina LAC recommends, among other options, spending at least $17 million to either retrofit or replace those machines with similarly 100% unverifiable ones. The LAC suggests replacing the 100% unverifiable paperless touch-screen systems with 100% unverifiable touch-screen systems that produce a so-called "Voter Verifiable Paper Audit Trail" or VVPAT.
Unfortunately, there is also no way to know after an election that any VVPAT has ever actually been verified by any voter, nor that it actually reflects any voter's actual intent.
Even more unfortunately, the LAC doesn't even refer to those VVPATs as "Voter Verifiable Paper Audit Trails" in their report. They incorrectly describe them as "Voter VeriFIED Paper Audit Trails," which is a blatant misnomer. I'll explain why in a moment.
Thankfully, the report also details the option of moving to an actual paper ballot system in SC. Hand-marked paper ballots are voter-verified by definition. When the voter fills in the oval next to a candidates name, or connects an arrow on the paper, we can reasonably know what their intent was (unless they misread the ballot, but that's their fault, not the voting system's.) How the vote on that paper ballot is then tallied --- publicly and reliable by hand, or by oft-failed, easily-manipulated optical-scan systems --- is a separate issue. But at least we can know who the voter intended to vote for, if anybody ever bothers to actually count the paper ballots that voters took time and care in casting.
While it's good news that SC's LAC has highlighted some new options for the state, it's a shame that they've chosen "truthiness", as Stephen Colbert might say, over the actual truth in attempting to inform state lawmakers of their options. In the very first sentence of the report, and throughout the rest of it, they completely misidentify what a VVPAT is by referring to it as a "Voter Verified Paper Audit Trail", which nobody can ever know if it actually is or isn't.
So the report ends up recommending, among other options, that the State Election Commission "upgrade" to equally 100% unverifiable touch-screen voting systems --- and then misdescribes that "upgrade" option, failing to advise lawmakers that the systems would still be 100% unverifiable even after that "upgrade".
'VerifIABLE' does not mean 'verifIED'
The SC LAC is not the first to make this seemingly small, but very important, mistake. But as the report they produced was in response to the former President Pro Tempore of the South Carolina Senate's request for had "a review of the voting machines used in South Carolina" because "He was concerned about the reliability of the machines and the lack of paper trail to confirm voting results," this is a point that the LAC should have taken very special care to get right.
Instead, they refer in the very first sentence of the 46-page report [PDF] to VVPATs --- the little piece of paper printed off by some touch-screen voting machines (though not the ones currently used across SC) --- as a "voter-verified paper audit trail". The third sentence of the report also refers, inaccurately, to it, in recommending that law makers might wish to "add a voter-verified paper audit trail" printer to their existing touch-screen machines.
Even the Executive Summary [PDF] for the document refers inaccurately to a "voter-verified paper audit trail (VVPAT)".
A VVPAT, however, is a voter-verifIABLE paper audit trail. There is no way to know after an election whether it has actually been verifIED by any voter. Using incorrect terminology to suggest to lawmakers that election results can be known to be any more accurate with the addition of this flawed technology is the opposite of helpful to them.
As we've explained many times over the years, studies by institutions such as CalTech and MIT have found that very few voters actually bother to check their printed VVPATs at the end of the voting process to make sure their vote has been printed on them accurately. Other studies, such as one done at Rice University, have found that even those very few who do bother to check the computer-created summary at the end of the voting process don't notice when the computer has flipped one or more of their votes.
A 2006 PBS News Hour report on voting machines some years ago showed voters not bothering to check their printed VVPATs after casting votes on similar machines in California (where most such systems were long ago banned for use in elections by all but disabled voters who may choose to use them.) There's even a moment in that report when a poll worker is seen specifically telling voters to not look at the VVPAT to verify their vote, that the paper trails are meant for administrative purposes only.
There is absolutely no way to guarantee that the results printed on a VVPAT produced by such machines --- even if anybody bothers to examine them after an election, and almost no jurisdiction in the nation ever does --- accurately reflect the intent of any voter.
Describing such records as "voter-verified", therefore, is a terrible misnomer. VVPATs are "voter-verifiable" at very best. So, during a post-election spot-check or "recount" of those VVPATs, should one ever happen, it is impossible to know if any voter has actually done so.
"I see your point," the LAC's Director Simpson told me after I explained the issue to him on Friday during a phone conversation. "That paper audit trail wouldn't necessarily reflect the way the voter voted if the machine messed up somehow and hadn't recorded the vote correctly on it."
"It might be helpful if I talk with the staff in a little more detail about this," he said, adding that he'd try to get back to me once he'd had the chance to "touch base with the manager who worked on the audit".
'Truthiness' still rules in South Carolina elections
When the CA Sec. of State carried out a landmark "Top-to-Bottom Review" of all electronic voting systems used in the state in 2007, one of their expert reviewers, the Computer Security Group at the University of California/Santa Barbara, created a video demonstrating how an election on a touch-screen system with a VVPAT could be gamed in such a way that even if anybody bothered to examine the paper trails produced by it during a post-election examination, they would still not be able to tell that the election had been gamed.
Simpson told me that his LAC report may have overlooked or misunderstood these points since the bigger issue, as far as they were concerned, was that SC's voting machines "have no paper trail at all". That is true, of course, but adding a VVPAT --- rather than moving to a system of hand-marked paper ballots --- wouldn't actually lead to voters knowing if their election results were any more accurate in SC than they are today.
South Carolina's LAC, whose motto is "Independence, Reliability, Integrity", has not produced a particularly reliable report. Given that "the purpose" of the LAC, as defined on the front page of their web site, is "to provide information which will assist the General Assembly and the public," the misleading information from their report is no small matter.
For example, in an article on the LAC report last week in South Carolina's The State, the paper quotes a number of legislators, both Democratic and Republican, calling for millions to be spent to add VVPATs to the existing voting systems, as recommended in the report.
"House Minority Leader Todd Rutherford, D-Richland, said the state should spend the money now to upgrade the existing machines," The State's Adam Beam reports.
"We want them so we will know our vote actually counts," a misled Rutherford is quoted telling the paper. "Knowing that people’s votes count is worth $17 million in an era when we have seen people’s votes not count."
Unfortunately for Rutherford, even if the state spends $17 million to add VVPATs to their currently unverifiable system, it will still be 100% unverifiable. The LAC report, unfortunately, doesn't make that clear and, in fact, misleads legislators to believe otherwise.
Frank Heindel, a tireless Republican Election Integrity advocate in South Carolina, tells me he finds the report's recommendations for "upgrading" the state's DREs to be absurd. "It would be like a doctor suggesting a 90-year old man have a facelift. It's a waste of money on cosmetics and would do nothing to improve his health."
Even after the national embarrassment of the Alvin Greene election in 2010 --- and even after other similar embarrassments in the state, such as the 2008 Republican Primary train-wreck in Horry County when the iVotronic touch-screens wouldn't boot up at all across the entire county, leaving people unable to vote and/or scrambling to find pieces of paper, and even paper towels, to try and use to cast a vote --- the state of South Carolina doesn't seem to "get it". The new report on their voting machines, as written, five years after Horry County and three years after Alvin Greene, is unlikely to be of much help.
[There are several other problems with the LAC's report that I may get to on another day, as well as some good things about it, such as noting that Gov. Nikki Haley has failed to remove or replace county election commissioners and voter registration board members "when they fail to comply with certification and training requirements" as per state law.]
Either way, even if a corrected report is issued by the LAC --- and, someday, South Carolina finally moves to "Democracy's Gold Standard", which includes hand-marked and publicly hand-counted paper ballots --- none of it will be done in time for the upcoming Colbert Busch contest against Sanford (or Bostic) which, according to a recent poll, is predicted to be a very close contest.
Stephen Colbert will take little comfort in knowing that, if his sister Elizabeth ends up being declared the "loser" on May 7th, she may or may not have been the one to have received fewer votes than her opponent. The results won't necessarily reflect the truth, as much as the "truthiness" unreliably reported by the 100% unverifiable voting machines still shamefully used in the great state of South Carolina.
UPDATE 4/2/13: Curtis Bostic just conceded to Mark Sanford in the GOP primary election run-off. So Elizabeth Colbert Busch will, indeed, face off with Sanford, who once held the 1st district seat, on May 7th. A new poll by Lake Research Partners [PDF], posted by the Colbert Busch campaign yesterday, shows her with a 47% to 44% lead over the formerly disgraced former Governor. That's almost almost the same as the 47% to 45% figure reported in a PPP survey in late March. With the run-off now over, we'll see if there is a "Colbert bump", or the opposite, in the days ahead, as the campaign for the seat --- held by the GOP since 1981 --- heats up.
Either way, as Vic Rawl --- who was inexplicably named the loser to Alvin Greene in 2010 for the Democratic U.S. Senate nomination --- reiterated to me in a discussion we had yesterday after I first published the article above, whatever the 100% unverifiable voting machines say on May 7th, and whether they are in line with pre-election polling or not, "the fact is, there's not a darn thing that anybody can do about it."
UPDATE 5/7/2013: Despite Colbert Busch's 9pt lead in the polls two weeks ago, and the race declared a "toss-up" by pollsters over the weekend, Sanford has now been declared the "winner" by a landslide tonight by SC's 100% unverifiable touch-screen voting system. Details now here...