For weeks, we've been reporting on the mess seen in the statewide "recount" of Wisconsin's very close and very contentious April 5th Supreme Court election between Republican incumbent Justice David Prosser and his challenger, Asst. Attorney General JoAnne Kloppenburg.
As The BRAD BLOG obtained evidence of new irregularities this week --- to add to previously reported revelations of, among other irregularities, ballot bags discovered "wide open" with mismatched or missing serial numbers as well asballots discovered completely unsecured, all in violation of the secure chain of custody, and other similar messes and mistallies across the state --- we wanted to find out if the state's chief election agency, the Government Accountability Board (G.A.B.), was able to confirm that the ballots counted during the "recount" were actually the ones cast on Election Day. And, if so, how they could confirm that.
To date, with about 25,000 "recounted" votes still to be canvassed from the now infamous Waukesha County, some 2,690 votes have been discovered during the "recount" to have been originally mistallied. That's according to the G.A.B.'s own figures, and includes hundreds of thousands of ballots which were simply re-tallied rather than counted by hand during the "recount," on the same computers which tallied them --- either correctly or incorrectly --- on Election Night.
31 out of the state's 72 counties counted some or most of their ballots by hand for the first time during the "recount" (including several of the largest counties). But, given the fact that supposedly "secured" ballot bags were discovered during the "recount" to have been "wide open" and not secured at all, with little or no explanation from officials, after being accessible for weeks by the same untrustworthy and partisan election officials, such as Waukesha's County's activist GOP County Clerk Kathy Nickolaus, who screwed up the tabulation in the first place, how is it that anybody knows if the ballots finally counted by human beings are actually genuine?
We've been trying, with no small amount of frustration, to get a definitive answer from the G.A.B. on that question for the past week, as well as from the campaigns of both candidates in the race.
The G.A.B. generally agrees with the campaign of the election's current leader, incumbent Justice David Prosser, that broken security seals, opened ballot bags --- even some that are torn open and taped back shut with duct tape, as seen in Waukesha County this week (photos below) --- and other violations of chain of custody are of little concern. So long as the newly tabulated results largely match the results printed on the poll tapes by the computer tabulators at the end of Election Night on April 5th, that is, essentially, close enough for government work.
The G.A.B. and the Prosser campaign, as we confirmed with each directly, and as the G.A.B. indicated on a recent post to their website published in the wake of the specific questions we've been asking them --- are placing their faith largely in the accuracy of the state's oft-failed, easily-manipulated, privately-manufactured electronic voting systems made by companies like Diebold, ES&S, and Sequoia. They are also placing faith in their election official colleagues around the state.
All of which begs the question: What's the point of having a "recount," or of using security procedures and physical seals for the ballots after the election, if violations of those procedures and seals are of little concern to the state's top election agency?
Worse, if the results printed on the poll tapes are the ultimate proof of the accuracy of results, what happens when --- as discovered among poll tapes from the City of Pewaukee in Waukesha County late last week --- the "recount" uncovers "Official Results Report" poll tapes dated a full seven days before the actual election was held?
Or, worse still, what happens when poll tapes failed to print at all on Election Day, as has been seen in a number of towns across the state?...
"Even if the container or [ballot] bag is somehow opened later, or if the chain of custody is broken," the G.A.B. wrote on their website in response to the concerns late last week, "election officials have the original print-out tape from the machine, as well as the electronic memory device from the machine. This enables election officials to determine the election night vote count."
Setting aside that the "recount" process in WI does not include examination of "the electronic memory device from the machine[s]" at all, the print-out tapes from the systems may enable election officials to know what the tabulation machines reported --- either accurately or inaccurately --- as the "election night vote count," but do those digital elements actually tell us what the intent of voters was? If they do, then why bother to have a "recount" at all?
"If the ballots had been tampered with between the election and the recount, there would be a break in the chain of custody and an unexplained difference in the results [of the 'recount']," says the G.A.B., suggesting that they seem to have little or no idea how election fraud may be carried out under their very own noses.
As reported here in great detail here over the past month, there were definitely violations and breaks in the chain of custody of thousands of ballots. If they had been gamed, in order to avoid getting caught in this very close, very high-stakes election --- for a 10-year term on the state Supreme Court during one of the most tumultuous periods in state history --- it's as simple as swapping in ballots to match the totals on the gamed poll tapes. (Gaming those tapes can be as simple as seen in this clip from HBO's Emmy-nominated documentary Hacking Democracy.) If one happened to have blank ballots and knew those blank ballots would not be reconciled, even during the "recount" (since they do not do so in WI), and one had possession of the ballots for a full three weeks prior to the "recount" --- as, say, Kathy Nickolaus or any other election insider in the state of Wisconsin had --- it would be worth both the effort and the risk to avoid detection if one had gamed the front end, as discovery of the dirty deed during the "recount" would otherwise result in some very hard time.
(Not that any really high-ranking election officials would ever do that sort of thing in election after election, right?)
But so long as the poll tapes generally match the "recounted" ballots, according to both the G.A.B. and the Prosser camp, there's no problem.
When we asked the Prosser campaign if they could confirm the authenticity of the ballots being counted in Waukesha and elsewhere where ballot bags were discovered with openings large enough to easily add or remove ballots, they had no concerns. Prosser spokesperson Brian J. Nemoir told The BRAD BLOG it all goes back to the machines which "spit out tape with a candidate total" on Election Night.
"You take those tape totals and compare them to the canvass totals," he explained, even though the canvass in WI doesn't actually include tallying of results, just a reconciliation of the number of ballots cast with the number of voters signed in to the poll books. "And then, finally, you have the recount. I think it's hard to conclude that anything was switched," says Nemoir.
Which Poll Tapes? The Ones Dated Seven Days Before the Election?
On Friday of last week, Barbara With, one of Kloppenburg's volunteer observers at the "recount" in Waukesha County, was startled to watch poll tapes being counted --- in this case, so-called "Voter Verified Paper Audit Trails" (VVPATs) printed out by Sequoia AVC Edge touch-screen machines from the City of Pewaukee --- even though the dates on them read March 30, 2011, a full seven days prior to the April 5th election.
The "Official Results Report" poll tape was dated 03/30/2011, with a time stamp of 1:40am. No one seemed to know why, according to With.
Top of the 'OFFICIAL RESULTS REPORT' poll tape, dated 3/30/11 at 1:40am, from the 4/5/11 Supreme Court election in Waukesha County's City of Pewaukee. (Photo: Barbara With)
"They confirmed the tape with the bad numbers was not a test," With told The BRAD BLOG. "The clerk [from the City of Pewaukee] said 'No, this is not a test. This is it. Who knows why the numbers are wrong?'"
Result section of the 'OFFICIAL RESULTS REPORT' poll tape, dated 3/30/11 at 1:40am, from the 4/5/11 Supreme Court election in Waukesha County's City of Peaukee. (Photo: Barbara With)
With says that when the clerk "was shown the piece of paper with the mismatched date, she swore, 'No, these votes were taken on April 5th, it's just a mismatched date in the machine.'"
Signatures on the 'OFFICIAL RESULTS REPORT' poll tape, dated 3/30/11 at 1:40am, from the 4/5/11 Supreme Court election in Waukesha County's City of Peaukee. (Photo: Barbara With)
One of the people overseeing the count on behalf of Waukesha County also instructed workers to proceed and count the results on the tapes after the questions arose about the date and time stamps. "I witnessed Barbara Hansen examine the tape to assure workers that they were not counting a test," says With.
The observer from the Prosser campaign was nonplussed.
"The Prosser guy was saying, 'Ya know when your computer blanks out and the clock resets?' and I was like, 'No, that's January 1st, 1980, not March 30, 2011!"
'OFFICIAL RESULTS REPORT' poll tape, dated 3/29/11 at 1:40am, from the 4/5/11 Supreme Court election in Waukesha County's City of Pewaukee. (Photo: Barbara With)
With, who had spent days as an observer at the "recount" in four other counties as well, couldn't recall how many of the mis-dated VVPAT votes were tallied for certain. "I don't remember, maybe 24, but I looked at every one of them," and they all had the wrong date, she says. She took a few blurry shots with her cell phone (as seen at right, and at the top of this article.)
Waukesha County, one of the most Republican-leaning in the state, hired Hansen, "a former deputy for the state Government Accountability Board and a 21-year employee of the state Elections Board," according to Lisa Sink at the Brookfield Patch, to assist retired Waukesha County Circuit Judge Robert Mawdsley in overseeing the count there.
Mawdsley had been appointed to do so by the Waukesha County Executive Committee after Nickolaus recused herself from the "recount" the day before it began on April 27, several weeks after the April 5 election, and a full week after Kloppenburg hadannounced her intention to file a "recount" and to request an independent investigation of Nickolaus. Kloppenburg charges that the G.A.B. was not suited to do the investigation themselves, given their close working relationship with Nickolaus and the other county clerks.
Despite her recusal, the ballots from the county of Waukesha have been in Nickolaus' custody since at least 4pm on the day following the election.
Kloppenburg had initially been in the lead on Election Night by a ridiculously slim 204 vote margin. At a press conference two days later, GOP activist and County Clerk Nickolaus stunned the state by announcing that, due to "human error," some 14,000 votes from the City of Brookfield had been left out of her Election Night totals,reversing the results and giving her former colleague and boss David Prosser a more-than-7,000 vote lead over Kloppenburg in the statewide race. Neither the County's Board of Canvassers nor the G.A.B. had been informed about the "human error" until the evening presser on Thursday, April 7, even though Nickolaus says she discovered it on the Wednesday morning after the election.
It wouldn't be the first time problems and irregularities plagued an election administered by Kathy Nickolaus. She has compiled a long and ignominious history of such "errors" since becoming County Clerk in 2002, after receiving immunity from criminal prosecution in exchange for cooperating with prosecutors during a scandal that sent a number of her and Prosser's colleagues from the Assembly Republican Caucus to jail.
Ballot bags holding those 14,000 votes from the City of Brookfield were among the ones found to feature some of the most disturbing irregularities during the "recount." While examining one of them before counting the ballots inside, Judge Mawdsley said it had "the widest gap we have seen in ANY of the bags so far," according to observer Mary Magnuson who took the extraordinary photos of "wide open" bags and scratched out serial numbers that we published here earlier this month.
The irregular City of Pewaukee poll tapes that With witnessed being tallied last Friday also included an "Official Zero Proof Report" poll tape dated the day prior to the "Official Results Report," on 03/29/11 at 11:01am. The "Zero Proof" is the test tape printed out on the morning of the election before polls open, to "prove" there are no votes already on the machines.
Polls in Wisconsin open at 7am and close at 8pm. Perhaps the date on the machines was off by seven days, for some reason, and the time by 4 hours. Perhaps the 3/29/11 "Zero Proof" at 11:01am was really printed on 4/5/11 at 7:01am as polls opened. If so, the "Official Results Report" dated 3/30/11 at 1:40am would actually have been printed on 4/5/11 at 9:40, a full hour and a half after the polls closed. That still seems curious, and begs the question as to why none of the signed witnesses on the tapes noticed the date/time discrepancies before or after the machine was put into service for the statewide election.
There is a form of "early voting" in WI, in that absentee voters may drop off their ballots at the municipal clerk's office prior to Election Day. Perhaps "maybe 24" of those absentee voters voted on a touch-screen machine at City Hall? But, if so, why would an "Official Results Report" be printed prior to Election Day?
With was troubled enough by the incident to write a short letter for Judge Mawdsley last Monday, which was also given to the Kloppenburg campaign and sent to the G.A.B., following what she witnessed the Friday before. The letter requested the Clerk from the City of Pewaukee "offer testimony into the record" concerning how date adjustments are made on the touch-screen voting machines. Her request was not fulfilled.
The Clerk from the City of Pewaukee, Kelly Tarczewski, has not responded to several requests from The BRAD BLOG, via phone and email, to comment on the irregularities. Neither has Deputy Clerk Ami Hurd. A perfectly reasonable explanation may exist for these irregularities, but as yet we cannot report what that may be.
Melissa Mulliken, Kloppenburg's campaign manager, says that while she hasn't been able to review the specifics of what happened in the minutes (they are not available to the campaigns or the public until after the count is complete), her general understanding was that those tapes were determined to be from pre-election tests, in contrast to the detailed information provided by With to The BRAD BLOG through several different phone conversations and emails.
After the counts in all counties are finally complete, the G.A.B. has three days to canvass and certify the results. Then the Kloppenburg campaign will have just five days to review all of the minutes and all of the results from all 72 counties, to determine if they wish to file a judicial review to contest the election and challenge the extraordinary number of irregularities that have been uncovered to date.
Or Maybe These Poll Tapes? The Ones That No Voter Ever Reviewed?
As if the opened ballot bags, missing and scratched-out serial numbers, and mis-dated poll tapes aren't enough, the minutes from the "recounts" in three different cities in two different counties have revealed that the so-called "Voter Verifiable Paper Audit Trails" on touch-screen voting machines actually didn't print at all, because the paper rolls were inserted backwards.
Most disturbingly, no one noticed or bothered to complain about it during the election! What does that tell us about the validity of the so-called "paper trails" printed out with touch-screen voting machines in Wisconsin, and the many other states which use the exact same unverifiable voting systems?
We are always told we can "trust" the results of 100% unverifiable touch-screen voting machines because voters review the "paper record" before they hit the "Cast Vote!" button. Unfortunately, as we've been explaining for years, no, they don't --- and here, once again, is more evidence.
The following descriptions are from the published minutes from "recounts" in several counties where they completed their work within the last two weeks.
From the minutes of Door County's Town of Forestville [PDF], where they use the Diebold AccuVote ES 2000 to tabulate paper ballots, and the Diebold AccuVote TSX touch-screen for disabilities voters and others who choose to use them:
In other words, it wasn't until the "recount" that anyone even noticed the "paper trails" or "paper records" of voter ballots didn't exist at all and therefore were never verified by voters. Officials simply used the touch-screens memory card to print them in order to count them at the "recount." They either accurately reflect the will of the voters who voted on April 5, or they do not. Nobody can ever know.
The same thing happened in two different towns, with different models of touch-screen machines made by a different manufacturer in Taylor County [PDF], where they use the ES&S Model 150 for optically-scanned ballots, and the infamously-failed ES&S iVotronic Direct Recording Electronic (DRE) system for touch-screen voters.
Taylor County, Town of Cleveland, Ward 1:
Taylor County, Town of Pershing, Ward 1:
The optically-scanned hand-marked paper ballots cast in Taylor were simply run through the same op-scanners again during the "recount," instead of being counted by hand.
When we called the Taylor County Clerk to ask for more information on the DRE "paper records" that didn't print on Election Day, we were only able to leave a message requesting a call back. We never received a return call. The nice woman who answered the phone at the Clerk's office and who had been at the count the day before, however, told us that "interestingly enough," the results of the recounts from the touch-screen systems were the most accurate of them all!
We took some time to explain why, that the "paper records" are simply a printout of what the computer says the votes are, whether that's what they really were or not. So reprinted "recount" results will always match the results printed on the results poll tapes on Election Night. Garbage in, garbage out, 100% unverifiable, faith-based voting. She had no idea.
'The Machines Will Save Us'
What a difference a few years make. Back after the 2004 Presidential Election, bloggers at the left-leaning Daily Kos website were permanently banned and their diaries purged if they dared discuss irregularities or the possibilities of fraud in that election. We've mentioned that depressing point a number of times over the years (most recently last summer during a radio appearance with Daily Kos founder Markos Moulitsas.)
Happily, things seem to have changed for the better over there --- at least if the great work and reporting from Wisconsin blogger "Giles Goat Boy" throughout the duration of the Supreme Court election "recount" is any indication.
In a detailed diary last weekend titled "The Machines Will Save Us", he observed some of the similar issues concerning the G.A.B.'s reliance on "poll tapes" as their "gold standard" for election results. He also looked at the conflict of interest for the state's G.A.B. in overseeing the election "recount" carried out by the very same clerks with whom they work closely, and to whom they offer training, throughout the year:
"Any fraud or incompetence the GAB might uncover among municipal and county clerks is an indictment of the GAB's ability to train and oversee those same clerks," he wrote. "The GAB has a huge incentive to avoid discovery of wrongdoing."
After 18 days of meticulously blogging each day's most noteworthy "recount" news, he lashed out at the G.A.B.'s continuing assertions that if everything matches up with the computer poll tapes, there is no cause for concern: "It's become a dog-and-pony show in many ways. 'Watch as we magically conjure up the same numbers we reported on April 5th with our infallible voting machines!'"
This story is too long already, so we'll not reprint it here, but Giles Goat Boy's closing rant of frustration in his article from Days 17 & 18 of the count is well worth the read.
The pseudonymous Daily Kos blogger also links to the same recent G.A.B. article we linked above, highlighting this particular assertion from the state's chief election agency:
True. Just because one can put his or her hand into a ballot bag and add or remove ballots, doesn't necessarily prove that someone committed fraud, any more than poll tapes said to be from Election Night, printed by computer systems prone to malfunction and malfeasance, prove that they did not.
As ballots in Wisconsin were not counted on Election Night, at the polls, by human beings, in front of the public, all parties, and video cameras, with results posted directly at each ward before ballots were moved anywhere --- as per "Democracy's Gold Standard" --- and because the chain of custody has been irretrievably violated in the case of thousands of ballots, we are left with nothing but guesswork as to the authenticity of the ballots and the true results of the election.
All we now know is that, in fact, criminals --- particularly those insiders with direct and largely unfettered access to both the tabulating machines and the ballots, peoplelike Kathy Nickolaus --- easily could have committed election fraud, as the supposedly secure, supposedly documented chain of custody, which is supposed to keep that from happening, has now been completely lost for thousands of ballots in a race said to have been decided by just a few thousand votes out of some 1.5 million cast --- a reported 0.488% margin between the two candidates.
Despite those chain of custody violations, and despite the ability to actually confirmthat all of the ballots counted during the "recount" were the ones actually cast, and despite the Kloppenburg campaign's objections to them, as noted for the record in the minutes when many of the irregularities were discovered, the ballots are being counted and included in the "recount" results as per WI law, at the approval and discretion of the Boards of Canvassers in Waukesha County and elsewhere.
But proving fraud is not the test for a post-recount appeal or overturning an election in the Badger State. The legal question concerns irregularities, according to Wisconsin statutes and case law.
While the statutes are very specific on which absentee ballots are to be removed from the count due to defects such as a missing witness signature (and, in fact, during the course of the "recount," defective absentee ballots were removed from the count for that reason), they do not speak to which defects and irregularities should disqualifynon-absentee ballots from being included in the count.
A review of the applicable statutes, however, makes it fairly clear that the question is not one of proving fraud, but rather, as explained in the footnotes to the Recount statutes (9.01):
"Irregularities," according to the statutes, are to be documented throughout the recount procedure, and we have reported on far too many of them over the past several weeks here.
The statutes explain that as the "recount" of ballots begins, "The board of canvassers shall then examine the container or bag containing the ballots to be certain it has not been tampered with, opened, or opened and resealed. Any irregularities or possible tampering with the container or bag shall be noted." [9.01(1)(b)3.]
As the count continues, "The board of canvassers or the chairperson or chairperson's designee shall make specific findings of fact with respect to any irregularity...discovered during the recount." [9.01(5)(a)]
Should the candidate choose to appeal at the end of the "recount," "The appeal shall be heard by a judge without a jury. ... Within the time ordered by the court, the appellant shall file a complaint enumerating with specificity every alleged irregularity, defect, mistake or fraud committed during the recount." [9.01(7)(b)]
That section of the code "constitutes the exclusive judicial remedy for testing the right to hold an elective office as the result of an alleged irregularity, defect or mistake committed during the voting or canvassing process." [9.01(11)]
And, as noted above, but worth repeating, "Generally, to successfully challenge an election, the challenger must show the probability of an altered outcome in the absence of the challenged irregularity."
That's just about it, according to the written statutes anyway.
In other words, as we read the applicable laws, the election could be overturned by the court if the Kloppenburg camp is able (and willing) to show that enough votes for David Prosser were "irregular" due to gross violations of chain of custody, wherenobody can swear to the authentic provenance of ballots being included in the count.
At this hour, with approximately 21,500 votes still being canvassed from the Waukesha "recount," there are 7,008 votes now said to be dividing the two candidates. If an appeal were held today, the Kloppenburg campaign would need to show that just 3,504 votes for Prosser, out of the 1.5 million cast in the election, are "irregular" or "defective" enough in some fashion that, if they were excluded from the results, there is a good "probability of an altered outcome" in the election.
Kloppenburg's campaign manager Mulliken insists they have made no determination about whether they will request a judicial review after the count is finally complete, but, she told us, "We are doing our best to do a complete and thorough job to object to these bags as they are counted, and reviewing that record will be our job in determining what we do after the election."
"We have to look at the record and see if it contains information and evidence that forms the basis for a claim. We don't know what we're going to do because the record's not complete," she said late this week during a phone conversation. "When the recount is over, we have five business days to decide if we're going to ask for judicial review. We're going to look at the facts and decide from there."
After nearly a week of repeated attempts at finding out how and if the G.A.B. wouldconfirm the ballots being counted were the same as those cast and, if so, how, spokesperson Reid Magney finally responded directly to the query, again, placing most of his faith in the machines.
"There are numerous security checks throughout the process to ensure the ballots being recounted are the same ones cast on Election Day," he wrote in an email. "There is a public test of the system before Election Day. Each ballot is initialed by two poll workers. There is a print-out made after the polls close. The hand-recount totals have been extremely close to the machine print-outs which, along with all the other protections in place, confirms that they are the same ballots."
As expected, the Prosser campaign, as they are still on the winning side of the reported results, have few concerns, and generally concur with Magney's assessment, even though they wouldn't confirm that they knew the ballots being counted were actually the ones cast. Spokesperson Brian Nemoir told us, "We are pretty confident in the results reported the night of the election and all three canvasses and now recount."
We suspect he'd be offering a different assessment if the numbers were reversed. Unlike Democrats (which Kloppenburg is not, she's independent) Republicans fight aggressively at any indication of fraud when they believe it might overturn a close election they appear to be losing. Unfortunately, they also have a tendency to invent such indications when needed.
Nonetheless, despite all that we have learned, Nemoir sees no problems. "I don't think the underlying story is that there are anomalies or concerns," he told us. "I think the underlying story is that this recount should provide great confidence. I think you've got a line of strong proof that should not invoke questions but should invoke confidence."
And if these new photos of duct taped ballot bags from the Village of Menomonee Falls in, you guessed it, Waukesha County (as taken by an observer who has asked not to be identified), doesn't "invoke confidence," we don't know what will...