Error? Or Fraud? (Could it be....SATAN?!?!?)

[farchivist.livejournal.com]

In the ongoing saga of the Wisconsin GOP/union battle, a new wrinkle is on the latest battlefield. This past week was the scene of the election of a Wisconsin Supreme Court judge. The incumbent is one David Prosser, a conservative who is generally considered to be aligned with Governor Walker. The challenger was the state Assistant Attorney General JoAnne Kloppenburg, noted state liberal. If Prosser loses his seat, this will tip the Wisconsin Supreme Court in favor of a liberal majority from its current conservative majority.


'Natch, in light of the Wisconsin union bill battle this Supreme Court election has become a proxy fight between the GOP and the Democrats. It is likely that whoever wins will be the decider of the fate of the union bill. In the days leading up to the election, there were interesting back and forth tidbits. Some examples:

- Prosser's campaign co-chairman, former Governor Patrick Lucey resigned from Prosser's campaign and endorsed JoAnne Kloppenburg, attributing his decision to Prosser's "disturbing distemper and lack of civility", while praising Kloppenburg for "adhering throughout the campaign to even-handedness and non-partisanship and exhibiting both promising judicial temperament and good grace, even in the heat of a fierce campaign."

- Politifact rates both Prosser and Kloppenburg as being a bit loosey-goosey with the truth, with Kloppenburg being more false.

- Prosser is reputed as being an intemperate man, but scuttlebutt is that it's a result of being ganged up on by court liberals

All the juicy tittle-tattle aside, the election took place on the 6th. Turnout was higher than expected, around 35%. The result was extremely narrow; Kloppenburg won the election by 204 votes. A recount, of course, was expected.

But.
BUT.
BUT.

There is a new wrinkle in the election. A Waukesha County Clerk by name of Kathy Nickolaus states that she made a mistake when entering votes into the system...and that there were an extra 7500 votes for Prosser, giving him the victory. Let's repeat that: 7500 votes for the incumbent, supposedly missed due to "human error" on the part of the Clerk, per the Clerk.

But it's just a mistake, right? Perhaps not. Some interesting things have come to light about Ms. Nickolaus. Some highlights:

- Concerns have been raised previously about her integrity in regards to vote counting. She's very secretive, having taken the election data collection and storage system off the county's computer network - and keep it on stand-alone personal computers accessible only in her office. Worse, she refuses to let the county IT department examine or support the system.

- Clerk Nickolaus says repeatedly that she imported the data into Access, but through a process of "human error", she "forgot to save". A fellow over at Daily Kos decided to experiment on his own. His result: Computer Says No.

- Nickolaus had worked in the Assembly Republican caucus during the time that Prosser served as the Assembly speaker. Supposedly, they had a very close relationship during the period. The caucus was one of four shut down following a criminal investigation into state staffers doing campaign work on state time and led to the resignations and criminal convictions of leaders in the Senate and Assembly for directing caucus and staff employees to engage in illegal political activity during their state employment. She was granted immunity during the investigation and resigned before the end.

A full analysis of the Waukesha vote problems are given here.

My opinion: Sounds a bit hinky to me. If it's an actual error, then I really question her ability to be in charge of elections. Her history does not put a good light on it and there's a lot of question as to why she waited 24 hours to make an announcement. However, only investigation will determine what the real deal is. I personally await the juicy gossip.

Hey, IT people, what do ya'll have to say about what Clerk Nickolaus says about Access? Can we get some opinions from all sides? Was the Daily Kos dude straight on?

Comments

[badlydrawnjeff.livejournal.com]

My opinion: Sounds a bit hinky to me. If it's an actual error, then I really question her ability to be in charge of elections. Her history does not put a good light on it and there's a lot of question as to why she waited 24 hours to make an announcement. However, only investigation will determine what the real deal is. I personally away the juicy gossip.

No, it doesn't sound good, but I don't know anyone actually involved, right or left, who doesn't see this as legitimate. A Democratic canvass watcher also confirmed the issue (http://dailycaller.com/2011/04/07/press-conference-prosser-picks-up-7k-votes-dem-canvass-watcher-confirms/) yesterday, and can verify the error. Considering this woman's previous lack of competence, and that the numbers coming out of the county now make more sense in terms of turnout and vote breakdown (Nate Silver here (https://twitter.com/#!/fivethirtyeight/status/56119393779847168) and here (https://twitter.com/#!/fivethirtyeight/status/56114655239741440) and here (https://twitter.com/#!/fivethirtyeight/status/56119574441107456)), my heavy skepticism when I first read it last night is tempered considerably. I wouldn't be surprised if something's not on the up-and-up here at this point, but I doubt it's the case given current evidence.

[farchivist.livejournal.com]

Fascinating. No doubt it will be investigated, I'm sure. I bet we can expect to see some hefty accusations of general incompetence. I'm also sure we'll see some litigation from somewhere on the election results, what with this foolishness.

I LOVE me a good dose of scandalous ambiguity.

[blueduck37.livejournal.com]

Regardless of whether fraud occurred here or not, can we all agree that partisan county clerks shouldn't be allowed to keep election results/data on their personal computers?

[dreadfulpenny81.livejournal.com]

Seems like Chicago-style politics are bleeding into Wisconsin. There were uncounted ballots in Mequon, WI that were allegedly shredded and thrown away. If that's true, it's illegal. Funny how the count for that area was highly in Prosser's favor...

[farchivist.livejournal.com]

From this pundit's article? (http://punditpress.blogspot.com/2011/04/wisconsin-city-caught-destroying.html)

[allhatnocattle.livejournal.com]

Should try out a good old paper ballot, the kind you mark an X on. It's simple to count and recount. And because there's so many returning officers involved in the tallying, it pretty impossible to fraud.

[kinvore.livejournal.com]

Yeah I wish we would adapt this system, I think they use it in Canada? At least I saw a story about it during the 2000 Presidential election.

[underlankers.livejournal.com]

Didn't we get George Bush by paper ballots?

[kinvore.livejournal.com]

Not sure if any of the links you provided mention it (that's a lot to go through haha) but what will make this look even more suspicious is that the Republicans needed over 7400ish votes to block an automatic recount. Two days later they find 7500.

It COULD be a coincidence but there needs to be a thorough independent investigation to make sure.

Not sure what to think myself at this point, I'd like to see more facts before deciding one way or another. Naturally being the partisan that I am I find it all suspicious but it could be that Prosser won fair and square.

[geezer-also.livejournal.com]

"I'd like to see more facts before deciding one way or another"

What kind of attitude is that in this age of instant communication? No one seems to need facts to form an opinion, go with your gut!!

[mahnmut.livejournal.com]

Hey? Cut? Yay! Thanks.

[airiefairie.livejournal.com]

I find your lack of lj-cut disturbing. =)

[chron-job.livejournal.com]

#1) Anyone who uses M$ Access for mission critical work product is an idiot. Well, let me soften that... they may not BE an idiot... but they are, at that moment, partaking of idiocy.

#2) Election results go a bit beyond the typical description of "mission critical" as used in a workplace setting. This is not a recipe collection. The fact that this clerk is a single point of failure (no backups, no access or oversight by other tech staff) is unconscionable. The fact that these issues came up in a previous audit, and had not been fixed is also unconscionable. This may be corruption, or it may be incompetence.

#3) The Daily Kos guy is right about what he says, as far as it goes... but there is plenty of room for human error in the importation of data from one source to another. "I forgot to save" MAY be a non-techie's plain English description for "I 'linked' a data table when I should have imported" or some such.

Bottom line is, a recount means recounting the votes... not re-tabulating someone else's count of the votes that they saved in access. Are there paper vote slips, or direct records of vote tallies from polling machines? Recounting those will ignore any kind of shenanigans that this county clerk may or may not be pulling.

[rasilio.livejournal.com]

"#1) Anyone who uses M$ Access for mission critical work product is an idiot. Well, let me soften that... they may not BE an idiot... but they are, at that moment, partaking of idiocy."

This.

Access is a great tool for what it is but it is not a real database and it is not suitable for enterprise level applications

[telemann.livejournal.com]

One state offical stated last night, he's never heard of anything like this, and at best, it's an example of incredible stupidity, and wondered why clerks in positions like this are elected in the first place.

[meus-ovatio.livejournal.com]

I thought it was strange that 14000 votes were found, giving someone a 7500 vote lead... which means the vote spread would be like 2:1 or something. If it was a 50/50ish split, the margins wouldn't have changed all that much. That's the only variance I see that is suspicious.

[badlydrawnjeff.livejournal.com]

If that county went 50/50, that would be more suspicious than these votes, which were in line with projections.

[the-rukh.livejournal.com]

Wait what, she took the vote count and put it on her private PC in her office and wouldn't let people access it? And then counted in word? What is this, how does even a preliminary vote count have any weight?

[underlankers.livejournal.com]

Frankly this is the kind of error/stupidity that becomes malicious even if not itself intended to be. With people now verging on cutting off pay to troops and the current political football nature of existing major issues that shouldn't be this, I think the Founding Fathers imitated the Roman Republic a little *too* well. This is just one example of that. If this is stupidity, it's a malicious stupidity. If it's blatant voter fraud she's lucky she did that for a Republican, not a Democrat. If it had been for a Progressive the screaming about the Hollywood-Soros-Liberal all-powerful Troika would have triggered a Zombie Apocalypse by now.

[rasilio.livejournal.com]

The really sad thing about this is that it should be relatively easy to design an electronic voting system that completely eliminates the possibility of errors like this while simultaneously providing 3 alternative methods of verifying the ballots.


It would look like this.

1) The Voter is issued a randomly assigned ballot number, however no one sees this number but the voter. The number is designed to identify the ballot location but nothing else beyond that.
2) Voter steps up to touch screen voting machine, enters his ballot number and makes their selections.
3) The Ballot is translated into a PDF document that displays only the votes cast (for example when voting on a ballot initiative if the voter selects Yes it will list that initiative followed by the word Yes, when voting for an office it will list the office and only the name of the candidate the voter selected). The PDF is displayed for the voter to review.
4) The voter can approve or reject the ballot, if they reject they are returned to the selection screen, if they approve the ballot image is stored in the machines memory and it is printed. On one side of the page is an exact copy of the Ballot PDF, on the other are barcodes or similar technologies that can read the actual votes cast.
5) The voter then reviews the printed PDF and in the touch screen can approve or reject it. If they reject there is a shredder and once the document is shredded the PDF is deleted and the voter returned to the selection screen.
6) The voter will continue the above process until they accept the printed image as matching their vote. Then they will place the ballot into an envelope such that the readable side is obscured but the barcode side is not.
7) The carry the envelope to the vote recording station where they are handed a tablet computer and their ballot is scanned in. The selections they made are displayed in the tablet computer, they can approve or reject the ballot again at this point. If they approve then the vote is immediately tabulated in the central database and again on a local database on the machine and the paper ballot is fed into a secure storage box. If they reject it the ballot is spit back out and they can return to the same touch screen machine, scan the ballot back in to retrieve their initial voting session and the ballot is moved into a different secure holding box. When they are returned to their voting session the electronic version of their previously printed ballot is marked invalid and not deleted but they are able to resume the process and produce a new ballot.

You now have at least 4 different methods of tabulating the votes.

You have the central database, you have the individual vote tabulating machine databases, you have the touch screen machine document repositories, and you have the physical paper ballots which because of the way they are printed leave absolutely no ambiguity as to the intent of the voter.

Could such a system be hacked to produce vote fraud? Sure, nothing is fool proof, however it would require a huge number of different people involved because there are so many checks and balances in place.

[allhatnocattle.livejournal.com]

Hmmm... add protections into a tamper-prone system sounds good on the surface. Now you suggest four levels of security. Why not 8? Or 12? or 40? How safe do you want to protect a system prone to tampering. You must work in IT.

Remember logging onto the internet for the first time? We used to worry about virus's then we had virus protection. Then after we had trojans, worms and such we protected against that too. But by now, with all the layers of various protection, the internet is safe as can be.

Lol, I prefer to trust my democracy to a system that's proven not to be prone to tampering.

[mrbogey.livejournal.com]

I think we can go a bit simpler and still be extremely secure. Giving the voter too much and having too many steps introduces the risk of human error gumming up the works. Ask any office drone how many tricks they take to bypass security practices that seem well meaning but are so onerous are complex that no one follows them... ie- complex building entrance rules end up with people stuffing a brick in a backdoor to keep it open.

Have a printer next to each touch screen voting booth that prints out an encrypted hash with a time stamp as part of the algorithm. The hash will contain all the encrypted info of who received a vote. That way if there is an accusation of hacking to change the inputted database than you'll have a paper record with an encrypted hash. Try hacking that. You'd have to do a man-in-the middle or some other hack that gets it before it gets added to the database.

But wait, there's more. That database since it has a time-stamp can tell if someone stacked the machine with votes before the election officially began or be checked to see if an oddity that indicates fraud such as a repetitive number of matching ballots cast in quick succession that would indicate someone standing there and "stuffing" the machine. All you'd have to do is feed the paper read-out into an OCR scanner and you should have a duplicate of the database.

That way a quick heuristic program could catch anything peculiar so investigators can deal with likely voter fraud quicker and more precise.

Of course there's still the possibility of a hack that cheats the input. You could always have the machine have a feedback page that takes the hash and converts it back to a ballot that a person can verify by looking at on the screen.

I think at that point, it'd be impossible short of having a widespread and open attempt at fraud to cheat the ballot box.

[gunslnger.livejournal.com]

The really sad thing about this is that it should be relatively easy to design an electronic voting system that completely eliminates the possibility of errors like this while simultaneously providing 3 alternative methods of verifying the ballots.

There is no way to design an electronic voting system that is secure from error of some kind. As I'm currently taking a computer security class from a professor who has done significant research into exactly that subject, I do have some basis for this conclusion. Your checks and balances only appear to be secure, I could go into how they could all be bypassed, but I don't think that's necessary at this point.

[gunslnger.livejournal.com]

Since the discrepancy was only in the unofficial totals and not the actual vote count, I don't see any issue, yet. Also, the article I saw earlier today, by chance, said that the count was off by 14,000-odd and it was off for both sides. So it's not as obvious and nefarious as you're making it sound.

Hey, IT people, what do ya'll have to say about what Clerk Nickolaus says about Access? Can we get some opinions from all sides? Was the Daily Kos dude straight on?

Depends on how it was being entered. Since it most likely was on a form, and not actually typing it into the database, it's possible. It would be a stupid design, but there's a lot worse running around. Read http://www.thedailywtf.com and see for yourself.