"Ohio Exit Polls not a 'Smoking Gun' for Fraud, Study Says"

[smartvoter]

I saw that we have threads on the study, but didn't see one showing how this study is being presented to the press. If it's a dupe, I apologize. The title of this thread is the press release title -- it's a ready-made argument.

http://www.votewatch.us/media/press_releases/aapor_2005

Ohio Exit Polls ‘Not a Smoking Gun’ for Fraud, Study Says

A just released analysis confirms pollster Warren Mitofsky’s assertion that the exit polls that put John Kerry ahead of George Bush in Ohio on Election Day 2004 do not necessarily indicate that there was fraud in the Ohio election.

Exit polls estimated that Senator John Kerry was leading for Ohio’s electoral votes, but not by a large enough margin to be called the winner. Had he won Ohio, he would have won the Presidency. However, the official result was a victory for President George W. Bush. The discrepancy between the polls and the results gave rise to widespread accusations of systematic election fraud.

The new study, commissioned by the Election Science Institute (ESI), was presented on Saturday at the annual conference of the American Association for Public Opinion Research in Miami. It looked at the results of the exit polls, which were conducted by Edison Media Research and Mitofsky International, and compared them to official results from 2004 and 2000.

The research team, led by Dr. Fritz Scheuren, used more detailed information from the exit polls than previous studies. The team was able to use this precinct-level information while preserving ballot secrecy at a local level.

“The more detailed information allowed us to see that voting patterns were consistent with past results and consistent with exit poll results across precincts. It looks more like Bush voters were refusing to participate and less like systematic fraud,” Dr. Scheuren said.

Dr. Scheuren is the current President of the American Statistical Association, and Vice President for Statistics at NORC, a research institute based at the University of Chicago.
Steven Hertzberg, project director at the Election Science Institute, spoke to the broader implications.

“We need to develop better tools to monitor our elections. The fact that there is debate over this at all shows that we need elections to be more transparent, more accountable, more auditable,” he said. “To increase public confidence in the system ESI has begun working with election officials in Ohio to help publish more timely election data so the public may verify for themselves that the voting and the counting is done accurately.”

The Election Science Institute (ESI) is a non-profit, non-partisan scientific organization based in San Francisco and founded in 2002 under the name Votewatch. ESI monitors public elections in the U.S. to identify voting anomalies which impact election results, and works with election officials to help them improve voting and election systems. ESI conducted its own exit polls in New Mexico for the purpose of assessing voters’ experiences.

[libhill]

1. Yeah, right

These pollsters are probably on the Repuke payroll.

[smartvoter]

3. They make the case for transparency in the body, but the headline is a

clear shot at US Count Votes.

[FreedomAngel82]

5. Wouldn't doubt it

So why did the same people who polled us who polled Ukraine got the winner right?

[smartvoter]

7. The part that bothers me is, when E/M made the RBR claim, why didn't they

step forward right then and say, "We saw that too. Here's what we saw in New Mexico."

Now, only after the E/M argument has been statistically slammed, they come forward and say they experienced the same?

[FogerRox]

12. yeah really--

ITs the UKRAINE stupid

[Selteri]

2. Again, they focus on one symptom

then go out of their way to ignore the other parts of the situation that also suggest there was fraud in this election.

Supressed black vote
Phone Calls telling people in Ohio to vote on Wednesday if they were democratic.
The private vote countings by Blackwell and others...
The fact that Blackwell RAN the OHIO campaign to reelect Bush...

The list keeps going, they just keep pointing to a symptom.

[smartvoter]

4. This is my biggest gripe. Exit polls are Addendum 142. nt

[On the Road]

8. That is Absolutely Correct

That is why it is important NOT to pound the exit poll argument unless it's demonstrably evidence of fraud.

All the other shenanigans are going down the rat hole -- voter suppression, lockdowns, individual vote-flipping, Diebold repairmen, etc. It's the same strategy that is constantly used by Republicans nowdays: debunk one argument, and everything else is assumed to be debunked as well.

That's why it's absolutely essential that any accusation be made to stick. And the exit polls argument doesn't fly.

[autorank]

11. Stop denying the obvious. Stop splitting the argument..borderline behavior

Exit Polls are an excellent pointer to election fraud. They are the type of circumstantial evidence that most prosecutors would kill for in any court case because they are empirical tests of actual events based on clearly defined laws and rules.

Here's the problem. There is no publicity about election fraud in CM (corporate media). Why the hell do you repeatedly insist on trashing exit polls when the people working on them are the last to say that they are the alpha and omega of election fraud.

This is a cheap rhetorical trick called a "straw man." You set up a bogus argument defined in an extremely distorted way (election polls distract from bla bla bla) and then you tear it down. By distorting the argument, you assure your counter argument...and it all goes to the point you want, which is to trash some of the best evidence we have.

People who really know statistics and how "games are fixed," i.e., the gaming, financial and academic communities, are well aware that the "fix was in" and they're quite comfortable with statistical analysis.

Mitofsky is the "Dog that didn't bark." He surely knows it was a fixed election yet he changed his weighting to fix his 2AM poll to support Bush. He goes around the country saying he did a bad job, including speeches and press releases. He's the pollster who didn't speak up, even for his own work.

Why don't you tell us what your real feelings are on election fraud?

Do you think Bush won?
Do you think Bush won Ohio?
Do you think it was OK for Richardson to ask $1.5 mil. for the NM recount?
Do you think that the tabulators were hacked?
Do you think that the VM companies should open up their source code?

I'd be quite interested.

[On the Road]

17. The Argument That the Exit Polls Demonstrate Fraud

Is based on one of several theories:

1) That precincts with electronic voting equipment were hacked.

If that were enough to swing Ohio, I would expect those precincts to stand out in the data as being anomalies. I have not seen any evidence that they do. Until that's shown, that's the strongest evidence against electronic fraud.

2) That the central tabulators were hacked.

If that happened in non-electronic precincts, I would expect the differences would be caught by the precinct-level people who reported the original results. If it were electronic machines only, see #1.

Maybe there's a way that the state-level totals could have been changed without detection, but I haven't seen it. My knowledge is only from local elections. Maybe there's something I'm missing that you can explain to me.

3) That the pattern of greater errors in competitive precincts indicates fraud.

This is what Elizabeth Liddle's effect addressed. The precinct results are consistent with her prediction and the effect needs to be removed to see if the remaining data suggest fraud.

Whichever theory you subscribe to, none of these arguments for fraud are persuasive, at least based on the data that I've seen. If they are, they have to be argued FROM the data based on the current state of the debate.

---------------------

As far as your questions go:

Do you think Bush won? As far as I can tell, Yes.

Do you think Bush won Ohio? As far as I can tell, Bush got more votes that were legally cast and counted. He might have lost if there were no voter suppression -- for example, if machines had been allocaed

Do you think it was OK for Richardson to ask $1.5 mil. for the NM recount? I think recounts are always good. $1.5

Do you think that the tabulators were hacked?I don't know. If they swung 130,000 votes in Ohio, I would expect the effects to show up in the exit polls -- namely that divergences would be systematically greater in districts with electronic voting machines than in others.

Do you think that the VM companies should open up their source code?Yes. Absolutely. I think it's an outrage that they're not. And I think Mitofsky should release his detailed data.

[autorank]

18. Bush didn't win.

The argument that the exit polls demonstrate fraud is based in the statistical analysis of the numbers. The explanations of fraud are multiple but you don't need to know how the crime was committed to know a crime was committed. The dead body, our democracy in the case, is evidence enough.

The exit polls were highly accurate in 1996 and 2000 which has been referenced all over DU. All of a sudden this year, they're completely off. I doubt it.

There are specific facts surrounding the behavior of EM that are of concern. Why did they revise the assumptions, add additional numbers and produce and entirely different result by 2am that matched the Bush electoral margin? This is bizarre to say the least. Why did they develop various rationales for what a lousy polling company they are to justify the mistake of the pre "cooked" 2am numbers? Why are EM people going around doing seminars and press releases on what a lousy polling company they are? This is unheard of in professional circles. Have you ever heard of a business advertising its incompetence? Why won't they open up their data files to a full forensic examination of just exactly what went on? And why are their virtual surrogates here at DU attacking TIA, sunshinekathy, etc. and their analysis of their EM's mis-analysis (I'm not talking about you)?

I'll tell you why, because they can't justify changing the 13000 Kerry-beats-Bush final at 10pm; and if they can't justify that, there are immediately questions, extremely serious questions, about the honesty of the results without any notion as to how the bogus results were achieved.

In court DNA evidence is based on probabilities suggesting a match between evidence with DNA on it and the suspect. In this instance, the Exit Poll analysis, exhaustively demonstrated all over Elections, is the electoral DNA evidence that demonstrates the probability that the reported vote is bogus due to a mismatch based on huge probabilities.

EM and the chorus of apologists are putting up smoke screens that fail to demonstrate anything other than the fix was in.

Open the files, allow a full examination, and, parallel to that, investigate to assure that those files have remained uncorrupted. Wouldn't want another fix again.

This is such huge evidence it forms a critical piece of the overall evidence of election fraud. Right now, it is the only piece of the evidence pool that demonstrates that Kerry won the electing. It also forms the basis for understanding the other elements of fraud related to machinery, hacking, etc. We ignore this gem at a great cost to our case. Read what's here and then read Jim Lampley's column and the opinion of the independent academics. Or just read TIA. That should do it.

[garybeck]

20. it's the scanners

you didn't offer an option on the scanners and punch card readers.

Need I mention TRIAD?

the recount was done illegally. the hand counts have never been done properly. IF (and that's a big IF) the ballots haven't been tampered with by now the answer to your question is in doing some manual checks against the machine counts.

pay close attention to Claremont county where recount observers swore under oath that they saw stickers covering up ovals that had Kerry votes colored in.

[Carolab]

21. I'm not a statistician, but this is weak.

"2) That the central tabulators were hacked.

If that happened in non-electronic precincts, I would expect the differences would be caught by the precinct-level people who reported the original results. If it were electronic machines only, see #1.

Maybe there's a way that the state-level totals could have been changed without detection, but I haven't seen it. My knowledge is only from local elections. Maybe there's something I'm missing that you can explain to me."

You admit that you don't know about anything except local elections. Yet, you surmise that precinct-level pollworkers would have caught discrepancies in the central tabulation numbers. Really? Then how come no one could count up the number of signatures in the pollbooks and match those to the results? How come there were "more votes than voters" in so many instances (whenever there was access to the data). And how come, if nothing "funny" was going on, did Triad go to such lengths with the local BoEs to ensure that there weren't actually random counts conducted and that the results would all match up "perfectly" (even if "cheat sheets" were needed), so that only ONE county got recounted in its entirety--and that one county showed a LOT of discrepancies, favoring KERRY. And, yes, there ARE ways that state level results can be changed without detection. It's been DEMONSTRATED how that can happen.

[On the Road]

23. To Make an Argument for Fraud

it's necessary to show how votes could have been changed across the board, under different voting systems, in wards controlled by Democrats and Republicans. And that it could have been done without anyone noticing or being able to detect it.

Electronic machines have this potential. Some of Bev Harris' early work was very good on showing how insecure the machines were and how they are easy to hack. It's not clear to me to it can be done across the board with different systems. It needs the same level of analysis.

As far as the changing the central tabulators go, checking should be very straightforward. The official results are widely available. If there are discrepancies, they need to be identified BEFORE the argument is made. There also needs to be a more robust theory of how statewide numbers could be changed across the board without detection. Without those things, it's not a convincing claim at this point.

--By "more votes than voters" are you referring to Cuyahoga County? That was the result of poor data presentation by the County -- they lumped absentee ballots for a larger area under a small number of precincts. I personally ran the totals and saw where they got the numbers. If there were other geographic areas you were referring to, I'd be very interested in hearing about them. But turnouts over 100% in Cuyahoga County should no longer be brought up as evidence for suspected fraud. It weakens the argument a lot to bring up discredited claims.

--I'm very interested in the reports of stickers placed over Kerry ballots. I didn't realize there were affidavits. This is the kind of smoking gun that could break the issue wide open. So why are specific provable issues like this taking a back seat to muddier statistical issues like exit polls?

--I think that some of the actions reported about Triad employees are extremely suspicious. I think the machines need to be audited and the employees put under oath to explain their actions. From what I've heard, much of it could be attributed to a "cover your ass" mentality in covering up shoddy machine operation. I don't think it's conclusive yet. But it's one of the best places to push.

----

At the end of the day, who cares what you or I are convinced of? The goal is to publicly uncover any types of fraud that were committed. Some ways of arguing are more effective in a public forum. Others aren't. It's very important to have robust theories and hard evidence before going public. The exit poll argument doesn't seem to me to qualify. Some of the smaller issues might qualify if they had enough prominence to get further legal action. Personally, I think the "16-million-to 1" and "C'mon, it's everywhere" arguments weaken the case for fraud and have caused it to be taken less seriously.

[Carolab]

26. What about the situation where the pollbooks were taken away

from the two investigators who said they saw these errors in them and were going to copy them but were not allowed to do so...then the BoE was left open all night and the pollbooks and ballots left out in the open. From what I recall, one of them (Katrina Sumner) was threatened. I believe I recall at least one other precinct in Ohio where the recount volunteers saw these discrepancies as well and were not allowed to more closely examine the books. Then there was the report (via Andy S.) of how a simple canvass of the precinct showed that the reported number of voters was impossible. You can read more about them at the various sites devoted to detailing these--including David Cobb's.

There were instances reported all over the country of "more votes than voters".

[On the Road]

29. I Think Any Kind of Behavior Like That

needs to be vigorously pursued. Blocking recounts and vote audits is an outrage. But OOH, we know from Bev Harris' torn trash bag stories that not all reports of suspicious activity are accurate. And sometimes the stories may be basically accurate, but there's less there than you might think. When charges of fraud go public, they have to be as bulletproof as possible. Newsweek, CBS, and others have learned that any questionable item, however small, can be used to cast doubt on much more solid facts and arguments.

I don't doubt, BTW, that fraud was committed, as well as legal and quasi-legal kinds of obstruction. Regardless of how many votes were affected, it ought to be pursued. But if 130,000 votes in Ohio were stolen, I would expect it to show up more clearly in the exit polls.

I'll check out the David Cobb site for more votes than people. I'm mostly familiar with the Cleveland controversy. (BTW, another example of how one false claim can overshadow others.)

Pursuing legal action this far after an election is tough, especially in Republican-controlled states. But the charges in Michigan against multiple-voting Decmorats may have a silver lining: it sets a precedent for charges against Republicans.

[mod mom]

24. May I add the faux homeland security alert in Warren County OH on election

Day?

Or what about allowing repug volunteers unsupervised access to unsecured ballots in Lucas County, OH?

Or what about the extra 3,893 bush votes at a Gahanna precinct (Franklin County OH) that appeared in a precinct that had only 600 registered voters and was at the HATE BASED mega church responsible for:

http://www.ohiorestorationproject.com/plan.php ? (*Pastor Rod Parsley has been examined in the latest Freepress.org article. Note: His precinct received some of the inner city machines whose absence caused the long lines. He is a Blackwell for Gov supporter...hmmmh sounds suspicious to me)

SO MANY QUESTIONS CERTAINLY DEMAND A FULL INVESTIGATION WITH SUBPOENA POWER, DOESN'T IT?

[m.standridge]

25. Well, and what about all those "dramatic shifts" in all those key

Senate races on election night?
Isn't it odd, that it all happened--all those "dramatic shifts"--at about the same time, around 1:30 a.m on election night.
I remember that Leslie Stahl was covering the Senate races. She made note of the probable emotions of the Democratic Senate race candidates in several states that had been thought "fairly safe" --all in the space of just a few minutes. Over and over, until her face took on a look, almost of surprise, as CBS was forced to "call" race after race, based on election returns that weren't seeming to jibe with some of the exit polls in THOSE races either.

Just so--at the same time, Ohio "dramtically" seemed to turn around, from a Kerry lead to a Bush lead.

Do you think those were all just a bunch of coincidences?
Coincidence, coincidence, coincidence.
Power failures in just those states that could undo things for Bush in the Electoral College.
Discrepancies in the exit polls in those states that were in question, that were back and forth.
Odd-appearing vote totals in rural versus urban precincts in Oklahoma.
Disputed county totals in northern Florida.
Odd computer tech behavior in key precincts and counties insofar as the Electors' votes.
A fishy "Homeland Security" alert in a SOUTHERN Ohio county that may have been getting a little close for comfort due to Edwards' place on the Democratic ticket.
In order to focus on the exit polls' inadequacies and shoot those down, and on that alone--you have to ignore all those other "coincidences".

[smartvoter]

15. I think it's a matter of balance. The exit polls are important, but just

one element.

[m.standridge]

22. REALLY! God--check out Arkansas this last time. Last-minute

Bush effort to stave off the votes that started pouring in when Clinton started campaigning here for Kerry. Didn't happen in 2000.
Clinton pulls the Democrats up in precisely those areas of the state where they are often weakest against the Bushs--i.e., the western part.
CBS tv affiliate KTHV here in LR on election day, had last-minute polls showing the race as "too close to call"--48-48%.
The fact is, THE EXIT POLLS EVEN MISSED SOMETHING. Those last-minute Kerry votes came in in the western part of the state, much more heavily than statewide, insofar as the change in percentages and turnout. The western and southwestern AR towns of Mena, Hope and Hot Springs were powerfully impacted by Clinton's eleventh-hour support of Kerry.
Yet, statewide, but externally-based exit polls, would have "missed" most of this, because they only "spot checked" those towns and that area, with no more coverage of them than anywhere else. And yet, with it that close, those towns made the difference. Again, THE EXIT POLLS MISSED SOMETHING, ALL RIGHT.
We had a power failure in Little Rock on election day. There was a power failure in Iowa. There was a power failure in Colorado--Denver County, no less. There was a power failure in New Orleans. There was a power failure in Iowa. There was a power failure in New Mexico. All those produced computer breakdowns/meltdowns that took hours to days to repair or deal with. Plenty of opportunities for friendly technicians to pull another Clermont County, OH thing. Then there was that bogus "voter registration" drive in Nevada, that ended up disenfranchising tens of thousands of NV voters, especially once state-level judges refused to act to re-enroll them on the rolls.
Do you think those were all just a bunch of coincidences?
AR + IA + NM plus what Kerry had, even without OH = 270 Electoral votes.
CO + NM + NV plus what Kerry had w/o OH= 271 Electoral votes.
Blow off Louisiana's New Orleans power failure (LA was hair-close in the last Senate race there--so much so that some demographers were projecting Louisiana as potentially "the next Florida", but blow it off.) There you are: THREE scenarios, Electoral scenarios, handily "fixed" by computer glitches and power failures.
Then, there's Ohio, of course, which, alone, without the others above, would have given Kerry the Electoral College.
And then, there's Florida--which, again, alone, would have put Kerry over. What have we even heard about Florida THIS time besides a lot of indication of Pinellas County, etc., vote irregularities?
Does this not sound like some kind of NASA-like "backup" thinking here?
As to the thinking to justify it: get the Popular vote closer or better than last time, then fewer will think about it. It also puts M/E on the spot--after all, if Bush did better THIS time in places like NM and IA, which were close before, and if Bush got a Popular vote lead THIS time, when he didn't before, doesn't that suggest E/M owes everybody a big apology for BOTH elections?
Hence, the apologetic, yucky E/M behavior.

[meganmonkey]

6. See this thread for info on ESI :

As you can see, stevenstevensteven, a.k.a. Steven Herzberg, posts here on occasion.

http://www.democraticunderground.com/discuss/duboard.php?az=show_topic&forum=203&topic_id=353045#353051

[organik]

9. Exit Polls were right.

Any time a study comes out supporting Mitofsky's assertions, it receives MSM coverage. Interesting. His assertions of what could have went wrong are ridiculous. He's grasping: poll workers too young...etc. etc. BULLSH@T. Doesn't the MOE allow for these things? When something is outside the MOE..consistently...for the same guy...something is wrong. It's called FRAUD.

Why has the work of USCountVotes not been covered in the MSM, or Conyer's work, or how about Jim Lampley's latest assertions.

I happen to still believe the exit polls were probably correct. Where's the raw exit poll data?

I hope TIA is working on a book, or documentary. Anybody know of any election fraud docs in the works?'

With all of the Bush admins. other horrendous actions (no time to list them all), how could one possibly believe they didn't steal the election?

[smartvoter]

10. You have to wonder why they won't release that detail... nt

[autorank]

13. smartvoter, You are too kind a person. To me, it's obvious!

They are afraid that they will be caught lying; they are compelled not to release it by "influence;" and they are whores. But that's just my opinion...

:hi:

[smartvoter]

14. It is to me, too. Trying to be coy... :-) nt

[autorank]

16. My sense of subtlty leaves me on this issue (haha).

:grr: ...breathe in, breathe out, let your muscles relax...:)

[smartvoter]

28. LOL -- I have those days too. nt

[TruthIsAll]

19. Gee, smart, thanks for spreading the news. Shows you are right on

top of things.

Guess its time to throw in the towel.
Mitofsky wins.

ESI is a mountain of integrity.
And Febbles Fancy Function will live forever.

Got any more news for us, smartvoter?

How does one become smart enough to prevent the touchscreen from turning Kerry votes to Bush votes? It happened 86 out of 88 times to people you and I never met. Does ESI know about it?

Do You?

Know this: The odds are 1 in 79 SEXTILLION.
Do you know how big that is?

1 in 79,000,000,000,000,000,000,000

Why don't you send that one over to MP?
I'm sure he'd be interested.

Oh, and there's a lot more hard evidence just like that.
It's not Faith-based Fancy Febble Function Fiction either.

Those are FACTS. Real ones.

Keep up the alerts.
As they say, keep your friends close...

[smartvoter]

27. Don't be a schmuck.

TIA, I have posted in support of you too many times to count. I posted this because it provides us with the exact language sent to the press (it is a press release).

I pointed this out when I posted it and make no apologies for doing so.

If we live in a box where we don't even read what the other side is saying, we will never break through.