OHIO: Shared machines at adjoining precincts caused voter problems

[sled]

Shared machines at adjoining precincts caused voter problems

http://www.cantonrep.com/index.php?Category=13&ID=197051&r=5

Friday, December 10, 2004

CLEVELAND (AP) — Voters using the wrong machines in adjoining
precincts at the same location apparently caused voting aberrations in
17 of 1,458 Cuyahoga County precincts, The Plain Dealer reported
Friday.

The paper said its review of voting patterns countywide in Cleveland
and its suburbs showed aberrations including a Democratic precinct in
which Democrat John Kerry got 334 votes and Libertarian Michael
Badnarik got 164. In the adjacent precinct, also at Benedictine High
School, Kerry got 299 votes and Constitution Party candidate Michael
Peroutka got 215 votes.

(SNIP)

The paper said unusual spurts in the Peroutka and Badnarik totals
apparently resulted from people in one precinct casting ballots just
steps away at machines meant for another precinct. The rotating
positions of candidate names on ballots meant people going to the
adjoining precinct had their votes counted for another candidate, the
paper said.

There was no way to tell whether Kerry’s or President Bush’s totals
were affected, the paper said. Kerry beat Bush 448,486 votes to
221,606 votes countywide, but lost Ohio by about 119,000 votes out of
5.6 million cast.

(SNIP)

CONTACT INFO: [e-mail]

[IndyOp]

1. Wooo hoooo!

I think this is one of Richard Hayes Phillips analyses in the Almighty AP! Maybe it was Juan Ganzalez from New York Post? Bless them for printing it, but geez I wish they would get away from the constant repitition of "no way to tell whether Kerry's or Bush's totals were affected" -- seems likely that they were, right?

The rest of the (short) article:

Together, Peroutka and Badnarik got about 0.5 percent of the vote countywide.

Katie Daley, an observer for the Democratic Party who spent Election Day at Benedictine, said voters waiting to cast ballots formed a single line between adjoining precincts and approached booths as they became available, without regard to precinct assignment.

Michael Vu, Cuyahoga County’s election director, said the findings would be investigated, and said they showed the need for more poll worker training.


:kick:

[EMunster]

2. This is what Arnebeck's talking about, I think...

[IndyOp]

4. This is part of what Arnebeck is talking about...

By the way -- I read what I think was your post on another thread that said to beware of anyone who offers their 'professional opinion' -- I agree and/but Richard Hayes Phillips, PhD seems to be a poet (from his website) -- not a lawyer or a scientist. I think some of his numbers are sound (some iffy) and all I can do is *HOPE* that Arnebeck and company have a witness who will stand up to a tough cross-examination. Then again, poets are pretty tough ducks.

:scared:

[eomer]

8. Richard Hayes Phillips is both a poet and a scientist

Excerpt from his resume:

DEGREES AWARDED:

Ph.D., Geomorphology, University of Oregon, Eugene, OR, 1987.

M.A., History, University of Oklahoma, Norman, OK, 1983.

M.A., Geography, University of Oklahoma, Norman, OK, 1982.

B.A., Politics and Geography (double major), State University of New York
at Potsdam, NY, 1979.

[IndyOp]

3. Phillips estimates 682 votes switched from Kerry to 3rd party in Cuyahoga

Richard Hayes Phillips, PhD estimates 682 votes were received by 3rd party candidates (Badnarik & Peroutka) that should have gone to Kerry.

Why should we believe that these 682 votes should go to Kerry on the recount instead of to Bush? In a state that has a Republican Governor the Republican candidate goes at the top of the ticket -- so this might mean that the order of candidates (Kerry, Badnarik, Peroutka) in different precincts would be rearranged but with Bush remaining at the top. (And, by the way, there is a large significant positive effect for any candidate at the 'top' of the ticket. Very undecided people often break for the person at th top).

STEALING VOTES IN CLEVELAND (NOVEMBER 18, 2004)
http://web.northnet.org/minstrel/cleveland.htm

[eomer]

6. In Ohio the candidate order is rotated

from one precinct to another so all candidates get their chance to be at the top an equal number of times.

[jmknapp]

37. Ballot order has no effect on candidate preference

For the presidential race on Nov. 2 anyway. Here's the average Kerry percent in all 1436 precincts, by each of the five possible ballot orders:

Ballot_order Kerry%
pbBKd 68.0
BKdpb 67.9
bBKdp 68.0
KdpbB 67.4
dpbBK 68.0

where
p Peroutka
b Badnarik
K Kerry
B Bush
d Disqualified

So there's practically no variation, and the variation that is there shows an opposite effect (i.e., Kerry did least well (67.4%) with the one ballot order that had him listed before Bush).

[mulethree]

42. jmknapp

an eye witness
Liberalnproud saw this happen

" Happened at the polling location I was watching also

There were two school districts in one precinct using the punch cards and they would swap out the "frame" depending on who voted. Backed up lines early. I think later on they just didn't even ask what precinct they were from which would make those votes invalid
"
http://www.democraticunderground.com/discuss/duboard.php?az=show_topic&forum=102&topic_id=1060300#1060307

If I read this correctly, two precincts shared a machine and it was sometimes re-configured for the voters precinct and sometimes not.
So the Ballots themselves would be screwed, so re-sorting them would not straighten it out :(

[jmknapp]

43. Precinct mark on ballot

I am still not clear though--is there or is there not some kind of mark or number on the punch cards that identifies the precinct that the voter is supposed to be in? In that case there might be impossible votes (such as for Disqualified) that would allow one to conclude the ballot was in the wrong machine (or had the wrong frame as you message puts it).

[Iceburg]

38. Not always ...

For example in Parma district:

At voter location ST. ANTHONY'S SCHOOL had two 2 voting precincts, both with the same ballot order (Bush 1st)

Prec Ballot Order
6A Bush Kerry DQ'd Peroutka Badnarik
9A Bush Kerry DQ'd Peroutka Badnarik

http://boe.cuyahogacounty.us/boe/ballots/PDF/PARMA06A.pdf
http://boe.cuyahogacounty.us/boe/ballots/PDF/PARMA09A.pdf
http://boe.cuyahogacounty.us/boe/PDF/votinglocations.pdf


At voter location RIDGE ROAD UNITED CHURCH OF CHRIST had two 2 voting precincts, both with the same ballot order (Bush 1st)
8A Bush Kerry DQ'd Peroutka Badnarik
8F Bush Kerry DQ'd Peroutka Badnarik
Ref:
http://boe.cuyahogacounty.us/boe/ballots/PDF/PARMA08A.pdf
http://boe.cuyahogacounty.us/boe/ballots/PDF/PARMA08F.pdf
http://boe.cuyahogacounty.us/boe/PDF/votinglocations.pdf

In Parma with 27 voting locations, the ballot order was repeated 14 times(all in locations with 5 or less precincts)

VoteLocName
--------------
THOREAU PK ELEMENTARY SCH
SCHAAF COMMUNITY CENTER
RIDGE ROAD UN CHURCH OF CH
PARMA SENIOR CITIZENS CENTER
RENWOOD SCHOOL
PLEASANTVIEW SCHOOL
PLEASANT VALLEY ELEM SCH
SHILOH JR. HIGH SCHOOL
NORMANDY SENIOR HIGH SCH
DENTZLER SCHOOL
DAG HAMMARSKJOLD ELEM SCH
MICHAEL A.RIES RINK
JOHN MUIR ELEMENTARY SCH
ST. ANTHONY'S SCHOOL

Question: Why was the ballot order repeated in some precincts at the same voting location and not in others?

[eomer]

40. The ballot order is rotated based on the master list of precincts

and the precincts that are grouped into a polling place are not always sequential on the master list.

Here's a thread that shows the rotation in Cuyahoga County:
http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x122913

So, I agree with you conclusion, but never said anything contrary to it.

[skids]

41. Good question.

Which is why we should be looking at whether the ballot order was allocated purposefully to damage Kerry.

If it turns out that an imporbable number of the ballot order changes were placed in heavy Democratic precincts (where most of the spoiled votes are Kerry) then we have raised some serious questions for the courts.

And, we have to look at them all, as well, to find those places where votes got swapped between Bush and Kerry but for some reason this did not raise too many 3rd party flags.

In addition, much of the data we have rolls Nader/disqualified votes into the overall undervote figure. So we really need a precinct-by-precinct full analysis of where the ballot orders were shuffled and whether they altered the outcome and who that favored.

Problems like this don't seem to be in the other counties as much, but we cannot rule them out. However, Cuyahoga needs to be done ASAP.

[eomer]

5. This states it differently than what I've seen before

Earlier reports speculated that the voter voted on the right machine and then deposited the ballot into the wrong box.

This one says that the voter voted on the wrong machine and then deposited the ballot into the right box.

The latter case is worse from the point of view of being able to count the vote because there may be argument on whether you can determine the intent of the voter. I wonder whether there's any precedent on this question.

In the former case, once you establish the voter's precinct (hopefully marked on the ballot) then there is no ambiguity in the voter's intent.

Also, in some cases there were more than two precincts in a polling place. Now it can be even more difficult in the latter case to determine intent because you have to figure out which one of several candidate orders the voter was presented with.

Edit to correct "former" to "latter" in the last paragraph.

[Mobrule]

7. Ohio

I'm just visiting here in OH (going to school) but it looks to me like the recount thing is going pretty well for the Dems.

[sabra]

9. Explain?

The recount is starting next week, as far as I know...

[EMunster]

10. The "aw shucks" tone of the "mistake" seems designed to undercut...

Arnebeck's charges in the plaindealer article:

http://www.cleveland.com/news/plaindealer/index.ssf?/base/cuyahoga/1102674912293811.xml

[eomer]

11. The plaindealer article also understates the problem

when it says:
<snip>
But the numbers appear small and the incidents relatively few.

Of 1,458 precincts in Cuyahoga County, 17 show what appear to be aberrations, with one of the low-profile candidates garnering at least 4 percent of the votes. In all, 942 votes were cast for Peroutka and Badnarik in those precincts.
<unsnip>

This ignores the possibility that in other precincts Kerry votes accidentally landed in the Disqualified position instead of Peroutka or Badnarik. Apparently the certified count threw away any punches for Disqualified so we don't know how many there may have been.

[Pacifist Patriot]

12. This is the point I've been trying to make with my media contacts

and the begging, pleading and shaming I've been sending to politicians. The size of the error is immaterial, it's the idea that no one can have confidence in the system with the "will my vote be tainted in some way?" fear. All of these "little errors" mean someone somewhere was disenfranchised. And that's unacceptable.

[eomer]

14. I agree completely

There are "errors" everywhere you look.

I'm at the point where I want someone to prove to me that the results were valid - not the other way around.

I want to know at every step why we should believe the results. Was the physical act of voting handled correctly, who had custody of the ballots, is there a reliable system for proving the chain of custody of the ballots, how do we know the punch card counter isn't hacked, how do we know the optical scanner isn't hacked, how do we know the central summary machine isn't hacked?

I want the recount to look at every one of those steps and prove to me one step at a time that the result is valid.

And that's in addition to the other techniques by which we know voter suppression occurred.

[ReneB]

23. but the "problem" is that if..

votes like in this case have switched from kerry to bush due the same "error / mistake" it is hard to find such an error, because bush gets votes, unlikely "third party candidates".

so what i say is: what happend in 17 precincts was only found because the "low-profile candidates" just got A LOT of votes which people afterwards investigated.. but if the error appears in a precinct where the rotation gives not an "low-profile candidates" but bush those votes, you will not find it.

sorry, i am german.. my english.. i hope you understood my point :]

[eomer]

26. Yes, I understand and agree with your point

and you're doing great with the English.

So, I agree that any Kerry votes that landed on Bush or Disqualified will likely go unchallenged in the recount.

I haven't looked into it but I wonder whether there are cases where the rest of the ballot could be conclusive proof that the ballot was inserted into a machine for the wrong precinct. For example, if two precincts at the same polling place have a different number of local candidates, local initiative questions or something like that then it may be possible to prove it was voted in the wrong machine. It seems more likely that this won't be the case and that it will be impossible to prove in those instances. Also, I imagine that even if it would be possible to prove in some cases, the recount won't look at the rest of the ballot once a clear vote for president has been recognized.

[ReneB]

28. so its not possible.

to track back because the ballots (punch cards) are not marked for the "right" machine?
thats bad..

[eomer]

32. Right

If the problem came about the way it's described in the articles then it won't be so easy as seeing that a card that's in the pile for precinct A is marked as being for precinct B.

You would have to look at other punches on the card to figure out that it was voted on the wrong machine or maybe you wouldn't be able to tell by any means.

I've heard that in some jurisdictions the machine is set up where you punch one hole that records which precinct the machine was configured for but as far as I know that wasn't done in Ohio.

[bones_7672]

13. From the Cleveland Plain Dealer - screw up by poll-workers

It would seem there was a screw-up by the poll workers at a few precincts.

http://www.cleveland.com/news/plaindealer/index.ssf?/base/cuyahoga/1102674912293811.xml

"But there are clear signs that, in some cases, poll workers erred in setting up the polling stations or misdirected voters. And voters, who often stood in long lines in cramped quarters, may have grabbed any open booth they could find, unaware there was a difference. In some cases, a combination of both factors conspired to produce bad votes."

[eomer]

15. Yes, I agree

that's what this one looks like.

But as I said in another post above, I'm skeptical about everything. I want proof.

Why do I want that? Because in 2000 the leaders of the Republican party, the SCOTUS and candidate Bush were all more interested in getting their man in than in determining who really won. In fact Gore actually won. So I've had it. This time I want proof of who won and I want it before anyone is inaugurated.

And, actually, to go a bit further than what I said above, I don't think the recount can possibly provide proof to my satisfaction. What I really want is a forensic examination of every step of the process for every different type of ballot. I don't think that's too much to ask given the fact that the wrong person is now the sitting President.

By the way, sorry to go off on this rant in response to your post. It wasn't anything you said - it's just something that's been building up for me and today I felt like spewing it out.

[jmknapp]

16. Critique of PD article

How does the PD reconcile these two statements:

"Voting book problems in the 17 precincts in question were easy to detect because the results for the third-party candidates were so atypical. Any mixed-book problems that may have scored extra votes for Kerry or President Bush are more difficult to detect because deviations from likely voter behavior are less obvious."

"But the numbers appear small and the incidents relatively few."

By their own admission they can't detect all the incidents, just the ones that raise a Badnarik/Peroutka red flag. So how do they conclude that the numbers are small?

Even as it is, probably thousands of misvotes are represented by the Badnarik/Peroutka cases. In the single-precinct polling places (where no mix-up was possible), Badnarik/Peroutka garnered 0.39% of the vote. In the multiple-precinct polling places, they got 0.57%. So 0.18% of the vote, or 1,200 votes were mis-cast.

Further, as the PD acknowledges, there are mixups that would not affect the third-party totals.

Some precincts are so heavily Kerry that Bush votes mapped to Badnarik/Peroutka would not be noticeable. Thus, for example, 100 Kerry votes could go to Bush, 5 Bush votes to Badnarik, and nothing would look particularly out of line.

Another possibility is, say, Kerry votes going to Bush and Bush votes going to Disqualified. Disqualified votes show up as undervotes and are not in the PD analysis.

All told. perhaps several thousand ballots at least in Cuyahoga County were switched.

The, "what, me worry?" tone is astounding. Here we see a system that has no checks against simply moving a stack of cards from one pile to another, causing erroneous (possibly maliciously so) results, and there is no call to immediately recount all cards based on their precinct stamps?

[Chili]

17. it IS astounding

I simply cannot fathom how the majority of "journalists" - both print and television - find it acceptable that thousands of votes were misrecorded, and still find no reason to investigate, in the most crucial state of the election, in the state that determined the election, in a state with an actively and openly hostile partisan Secretary of State who also served as co-chair of Bush's re-election campaign. I mean... it truly boggles.

And they wonder why people think there's a "conspiracy" silence within the media? How in the hell can we NOT think it?

(PS: Joe, your stuff is GREAT. I'm still in the middle of that project, deadline stifling my ability to work on anything here beyond just keeping up with the articles... I'm so glad you're still on it :hi: )

[eomer]

18. I'm also concerned about the possibility of moving cards

from one stack to another. But I'm hoping that there is a way to investigate that.

You wrote:
>>Here we see a system that has no checks against simply moving a stack of cards from one pile to another...

Isn't there supposed to be some check against that happening? Aren't the ballots supposed to stay locked in a box and then isn't any handling of the ballots supposed to be done with witnesses from both parties, with signatures, etc? I hope there is and I want a forensic investigation into the chain of custody of punch card ballots.

There was another thread this morning saying that the wide range of irregularities that have been discovered, along with the exit poll discrepancy, is justification in court for broad discovery covering all these issues. I sure hope so.

[jmknapp]

20. Yes, a hand recount would find at least some errant ballots

That's why I can't believe that the bottom line of the Plain Dealer article is:




"What, me worry?"

[eomer]

21. LOL (n/t)

[eomer]

22. Seriously though,

I'm concerned about the details of the recount.

Even a hand recount won't discover the case of a Kerry vote landing on the Bush slot. Not, that is, unless you can tell from the rest of the punch card that the ballot *could not* have been voted in the proper machine for the precinct where it's being counted.

That's why I'm saying we need a forensic examination. I'm afraid that the recount procedure will gloss over and ignore discrepancies that could be proven if you actually looked for them.

[ReneB]

24. so the punch cards

are not marked for "machine #1" or "machine #2" in the same precinct?
means, a handrecount would give the same result? just because "ballots" that were for "machine #1" are now mixed with the ballots for "machine #2" ?

[eomer]

29. The punch cards aren't marked for a particular machine

they are marked for a particular precinct. All machines for a precinct should be identical.

When I said that a ballot was voted on the wrong machine I meant on a machine from the wrong precinct.

[jmknapp]

25. Forensics

Not, that is, unless you can tell from the rest of the punch card that the ballot *could not* have been voted in the proper machine for the precinct where it's being counted.

That is a key point. I think forensics could tell a lot. Since downballot races differed between precincts, there may be punchs that uniquely identify the machine the card was punched on.

The prime example is "Disqualified." A punch in that column is a prima facie case that the card was punched on a different machine. In most precincts, the machine it was punched on will be obvious.

Then as I say the downballot races could be used to corroborate.

I have a feeling (hope) that this is going to be done eventually anyway, if not by Blackwell, then by independent organizations under FOIA, and the results should be interesting.

[ReneB]

27. i agree but..

forensics wouldnt that take "ALOT" of time?
only solution would be a revote in the counties where the problem appeared if you ask me

[jmknapp]

30. Not a whole lot I wouldn't think

The punch card can be read on any IBM-type card reader. Each polling place only has a limited number of possibilities on ballot order and the criteria to differentiate and sort them might be straightforward.

It will take longer than the time left before the inauguration, but IMO the chances of having a new election or stopping W's inauguration are nil. So it's a bit of an academic exercise from my point of view, along the lines of the analysis of the Florida ballots by news organizations in 2000.

[eomer]

36. It shouldn't take very long

if access to the ballot cards is available.

I haven't given up yet on seeing the candidate that won this election be the one that gets inaugurated. I think one instance of that is already one more than enough. Whichever candidate really won, that's who I want inaugurated.

[eomer]

34. I'm sure it will be done eventually

I'm just hoping it won't come about the way it did in 2000 that it is done eventually, after the inauguration, and then nothing will be done about it.

[Hamoth]

19. It's called an editor

You asked why they have such a rediculous contradiction in teir article. I beleive that many journalists are beginging to find out how bad this is, and see the big horrific picutre. They write a report that is probably even more scathing that what gets published, their editors then balk at the idea of going "out on a limb" and calling the shrub on this. They then insert these little lines "Though it can't overturn the election..." and "Incidents are few"...

I wish I had saved the link...i haven't been able to find it again. Nov 6th, there was an article I found through google news titled "Despite violence, election goes smoothly." I felt that really sumarized the divide between concerned members of the press and the conservative editorial staff. All articles about the election were given a nerf padding before being released.

[pgh_dem]

33. disqualified votes

I'm kinda having a tough time understanding why there is a 'disqualified' option anyway.
And if there is some bizarre necessity, this definitely should not be rotated, instead be the last option like a 'none of the above' vote...not floating around in the middle of valid votes.

[jmknapp]

35. Disqualified=Nader

Nadir was disqualified as a candidate too late to be removed from the ballot. That's the story. You're right it is confusing. On my own (Danaher) electronic machine in Franklin County, there was a name at the top (Badnarik or something equally obtuse to me at the time), followed by a blank spot (they just put a sticker over Nadir), followed by the rest of the candidates. It took me maybe 15 seconds to figure out what the hell was going on. "Badnarik, who, what?"

The weird thing is on these Danaher machines it was evidently possible to still push the button under the blank spot and record a vote for Disqualified.

Since Nadir was removed so late, the Disqualified spot had to rotate just like his name had.

On the other hand, it makes identification of misplaced punchcards somewhat easier.

[pgh_dem]

39. good point

Might be able to prove the intent of the voter if they voted on the wrong precinct machine (iow the pattern of votes would match up with all democrats or whatever on the assumed correct precint guide, but with the right precinct card in the wrong machine, then cast in the right box, you could have a totally crazy set of votes). I assume that the other miscount problem would be corrected by properly sorting the cards by the right precinct, if they voted on the right machine, but somehow ended up in the wrong precinct's box.

[skids]

44. Unofrtunatley we have to hunt them out.

A lot of counties are not providing separate numbers for disqualified/Nader votes. So all we have to go on is the undervote.

It would be easier to tackle this from the other end -- distill a list of precincts, polling places, and which ballot rotation they had.

[jmknapp]

45. But the DQ votes can be read with proper software

The holes are in the cards. It's the software that's ignoring them.

I have a spreadsheet with all the ballot orders, polling places, results and so on:

http://copperas.com/cuyahoga/cuyahogashuffle.zip

Joe

[november3rd]

31. Stealing the election

Anybody who tries to bring this up in a court challenge or other legitimate, legal public forum is going to get shouted down by the Republicans for, "Trying to STEAL the election."