Before, during and after last week’s historic Wisconsin recall elections, we detailed why there is reason to be concerned about the computer-reported results of the votes cast across the state, even with the majority of those votes having been cast on paper ballots.
In short, the results are, and will remain, 100% unverified unless someone bothers to get a court order.
As we previously reported…
In Palm Beach County, FL, the failure was discovered during a state-mandated, post-election spot check of 2% of the paper ballots. In New York City, it took nearly two years before the failures were discovered after the New York Daily News was able to examine the paper ballots via a public records request. In Oakland County, MI, election officials were lucky enough to discover the failure during pre-election testing. And in Leon County, FL, the hacker — a computer security expert — revealed the op-scan system flaw he exploited to flip the results of the election in an Emmy-nominated HBO documentary.
In all cases, it was only a hand examination of the paper ballots that revealed the mistabulations by the op-scanners.
Unfortunately, in Wisconsin, no such post-election hand-examination takes place for any ballots — at least not without both a recount request and an order from a judge. There are no spot-checks to ensure accuracy, so there is no way to know if the results reported by the computers reflect what the actual voters said in Wisconsin. The reported results are therefore either accurate or they are not. Who knows?
We explained why all of the above means that we cannot know if Republican Scott Walker actually won his gubernatorial recall election anymore than we can know if John Lehman (D) won his challenge against incumbent state Sen. Van Wanggaard (R) in the District 21 state Senate recall on the same day. (If Lehman is certified as the winner, the Democrats will take majority control of the state Senate. Wanggaard should demand a hand count of all ballots. To do so, he needs to file for a recount, and then beg a judge to allow that count to be done by hand, rather by the same machines once again, as is the standard under the WI election code.)
For those who think such concerns are overwrought, if the above instance of op-scan machine failure aren’t enough for you, take a look at what happened last week in three different counties in Montana, on the very same day of the recall elections in Wisconsin…
This is from Keila Szpaller’s report last week in the Missoulian…
Leigh Riggleman pulled an all-nighter in Lincoln County after running into problems with a ballot scanner in the primary election.
“There was no fix,†said Riggleman, assistant election administrator, on Wednesday. “The technician adjusted and adjusted and adjusted, and called, and talked to people, and it was merely a matter in some cases of running them (ballots) through a second time.â€
At least three counties reported problems with ES&S 650 ballot scanners in this week’s primary: Lincoln, Sanders and Powell counties.
…
“They finally determined that the ovals were a little bit too dark, and the scanner was just finicky enough that it didn’t like the darkness of those ovals,†Riggleman said.
In some cases, the county went through an official duplication process for ballots that just wouldn’t count, she said. Riggleman said she hasn’t run into trouble before and hopes not to in November.
“My understanding of the problem is that it can be taken care of in the next election by the printer not printing as dark an oval as he’d printed this time,†she said.
So now we have an issue with the ink used on the printed paper ballots themselves causing false-positives for the optical-scanner. In the instances mentioned above, fortunately, the machines issued error messages. How many machines do not, and simply presume overvotes, or record a candidate who the voter didn’t vote for as the selection for that voter?
This would seem to be yet another new vector for the malicious gaming of elections. Use slightly darker ink on one candidate than another and you force a vote for him/her, or otherwise force an overvote.
ES&S claims to be the largest supplier of voting systems in the nation. Their oft-failed machines — we have lost track of how many election results their systems have lost, thrown or otherwise just completely screwed up over our past near-decade of covering this stuff — will tally some 50% of the country’s ballots in this November’s Presidential Election.
Making matters even more troubling, the “fix” in this instance, as noted above, was for election officials to “remake” the ballots in secret, instead of simply counting those ballots publicly by hand. Hopefully they “remade” the ballots accurately, as per the voters’ intent, but, again, who knows? We’ve previously written about the extraordinarily offensive practice of “remaking” ballots, as is done in Wisconsin and in many counties across the country when, for any number of reasons, paper ballots are not being read by the optical-scan computer tabulators.
Szpaller’s piece goes on to explain how, despite the problems encountered with these systems, election officials simply cannot afford to replace them at the outrageously inflated cost of $50,000 to $60,000 each!
“I’ll tell you what,” Riggleman, Lincoln County’s election administrator is quoted as saying, “my ballot scanner is only about four years old. It should not be expired.”
On the other hand, citizen eye-balls used for public, precinct-based hand-counting, which is actually overseeable (as per Democracy’s Gold Standard), never expire and, unlike computer tabulators, they cost next to nothing each year.
In Sanders County, where the problem also occurred, the Missoulian continues…
…
Eventually, she said counties around Montana may need to budget for updated machines, and that will be a large expense for everyone.
And then there Montana’s Sec. of State who had heard nothin’ about nothin’ despite the all-nighters pulled in several of her counties the day before…
Accurate results? Who knows?
Who, other than those of us here at The BRAD BLOG, seem to actually care anymore?
“to dark” How? exactly?
Is the border too thick and thus the pre-printed impinges on the area to be reserved for marking?
Is the ink in the boarder too opaque in the infra-red spectrum detected by the scanner?
What exactly does “too dark” mean here? Can this be replicated? You like science instead of as a post-hoc myth of how the scanner screwed up.
If you want ot examine the ballots to see if the inks have different darknesses in the IR spectrum and the visible spectrum the montana officials should view the ballots with this IR re-enabled Sony HandyCam.
The reason I bring this up is that most “optical” scanners used in election are actually Infrared scanners. This is why red ink is such a problem for scanners. It reflects infrared almost as much as unmarked paper.
I have always thought a cool test would be to pre-paint an IR ballots with an IR opaque, optically transparent security ink, and see what the scanner registers. The super dark (in IR) security ink or the pen, or the pencil mark.
G-d forbid we test voting machinery to understand what kinds of grifts and cheats are possible.
Brad, would you clarify something for me? When kids take the SAT or ACT tests for college, they are always optically scanned using the same fill in the bubble format. So far as I know, they are considered reliable and presumably inexpensive. Why are the balloting machines so eminently screwy and expensive?
Unverifiable electronic voting is the bane of democracy. It’s been proven (by a Republican no less) that the Ohio vote in 2004 was rigged resulting in thousands of Kerry votes being flipped to Bush votes. I’ll give you better than even odds that Walkers recall vote, as well as the results of the recent Wisconsin State Supreme Court elections were rigged as well. These people are playing for keeps and refuse to be daunted by something as trivial as democracy.
Glidwrith
Easy answer – the folks involved with the test, its scoring, and the way it is used have no incentive to fake the results and kids actually can (and regularly do) take the test over as many times as needed to get a result that reflects their abilities and aspirations.
I have never thought of your point and it is interesting. In the end, I imagine that what technology is reliably capable of doing and how people use it wholly-depends on what they want to do.
In the case of elections, repugs and the economic/political interests they represent want to STEAL elections.
If we are going to have a rational dialog in this country, all questions about the integrity of elections need to be resolved; hence even if we can theoretically use these machines in a more responsible manner, there is a great amount of baggage here and always the opportunity to slip back into black box voting hacks and elections that are not accurately or properly verified.
But who has the funds for the lawyers to even ask for the verification count? The Dem’s or the unions? As an individual, I would count ballots until it is done accurately. I don’t know how to get the process started.
People care. Don’t underestimate the spreading effect on the subconscious of websites like yours.
@Glidwrith (#2)
Valid question, I’d say. If the SAT’s etc were so error prone, folks would be screamin’ to high heaven. How can their machines work so well and the opscan voting machines work so horribly? And who’s to say the SAT results aren’t hacked or error prone? Who would know? Do they have spot checks? Do they tell anyone if they find errors?
How about the opscan systems for tests in large universities? What percent error there? Couldn’t enterprising students hack those results for a better grade? Has it ever been done?
I believe that the machines that are used for voting were deliberately designed to flip votes or whatever other tricks the ‘owners’ have up their sleeves. We have a sizeable number of people sitting in Congress who went into the election behind by 5% or more only to appear on the morning after winning by 5% or more! This is no accident. Saxby Chambliss of Georgia is one of them. Remember the filthy campaign he ran against Max Clelland the Vietnam Vet who lost three limbs in that useless war? He was one of those who was trailing by a fairly large percentage only to wake up the next day winning by about the same amount. We are fffffed people unless we get rid of those damn machines. There is no reason on earth why we cannot have paper ballots which could be handcounted at the voting precinct within a couple hours of the polls closing.
@Lara #7: also very good questions. Maybe another way to get at this problem is to call into question the SAT machine results or determine their error rate and accuracy. We win either by showing those machines are inaccurate (which means the students and their parents will scream bloody murder for accurate results which will spill over into the balloting machines) or by getting a clean standard by which to judge the balloting machines and asking the very hard question of why they are compromised in the first place.
And how can Germany get by without electronic voting? Maybe because they’re a real democracy.
There must be some DoJ or other investigative agency that is following all of these incidents including the incredibly scary notion that WI has election results that are tabulated off site. We are reading about machine errors and reported machine vote rigging for 10 years now and all that has happened is that the companies are all under the same Dominion banner now?