New article on exit polls

[garybeck]

Hi friends,

I weaved some of TIA's numbers and tables into an article that is going out later today.

For most DUers there won't be anything new in this article.

However I felt it important to write this article for the following reasons:

-TIA's research needed to be published outside of DU

-I wanted to put out something short, in the form of an article as opposed to a report. To hopefully reach out to the more casual and uuninformed reader (people we need on board).

-Reports on exit polls have shown there is a problem and usually call for an investigation but to not mention anything about the explanation - voter fraud, electronic voting machines, etc.

Here is the article:

http://www.solarbus.org/stealyourelection/articles/0130-exitpolls.html

Exit Polls Made Easy
Something's fishy, and it ain't just Ohio

by Gary Beckwith
January 30, 2005

It's been said that exit polls are unreliable and that they can be interpreted many ways. But a look at the numbers provides some solid information anyone can understand, that is beyond interpretation.

These three facts can easily be derived from the exit poll data:

1) In 43 out of 51 states (inculding DC), the "official" vote totals differ from the exit polls in the direction favoring Bush.


2) Bush's victory exceeds the mathematical margin of error in 15 states. Kerry's victory does not exceed the margin of error in any state.


3) Exit polls were accurate in 2000.


In short, the gap between the exit polls and official vote (nearly all in the direction favoring Bush) is extremely large. Just how large?

The question that statisticians would ask is, "What is the probability that this could happen by random chance?"

Inquiring minds would like to know, and it's not just statisticians. It's the patriotic folks who care about the state of democracy in the United States of America.

<snip>.... rest of article is here:

http://www.solarbus.org/stealyourelection/articles/0130-exitpolls.html

peace,
gary

[EFerrari]

1. Thank you, Gary

Every time I feel too tired to even think about this any more, one of you great guys show me I'm not and I can. :)

[Wilms]

2. Smoked Fish

Good one.

A few ?'s.

You wrote, "Approximately 30% of the votes cast in the last election vanished into thin air the moment each ballot was cast..."

Is that right?

Am I (reasonably) correct in assuming that that getting an acurrate recount from a DRE/BBV is improbable?

Considering the Voting Systems employed, the amount of votes we might hope to recover, which states would you guess to be the best to recount? (We'll leave aside, for the moment, legal obstacles and additional fraud marring a recount.)

How many "verifiable" votes might we hope to garner.?

[garybeck]

3. my sources tell me

there's a 11/3 USA Today article that states that 28.9% of all votes were on DREs. The article says source: Election Data Systems.

Theoretically, all the rest had some form of a paper ballot that could be recounted.

I don't know which states should be recounted. I've lost faith in recounts. I guess I'd start with the states that were far off their exit polls and don't have a lot of DREs.

Gary

[Wilms]

5. I see.

I misunderstood the term "vanishes".

The reason for the other questions are not that I have faith in the real-world recount process. Perhaps, the opposite.

For a reality check, I'm hypothesizing Kerry having contested.

He would have gone after Ohio or Florida or even a combo of other states to secure electoral victory. But then he'd still have a popular vote deficet. So off to other states in hope of recovering those votes. I'm trying to get a sense of such a scenario.

[garybeck]

9. It's too late now but I would have seen something like this

One or two states are recounted, and errors are found.

Then because Diebold, or ESS were involved in the problems, all other states who used their equipment would be recounted as well, and the whole "popular vote" thing crumbles.

Gary

[Stevepol]

11. About 80% of the vote was counted on DREs. The 30% figure

applies to touchscreens. I don't know whether that means that 30% of the DREs are touchscreens or that 30% of the vote overall is counted on touchscreens. I think the former. Either touchscreens or optical scanners or any of the other DREs can cheat if the programming is made that way. Don't know where USA Today got their figures. I've seen the 80% / 30% figures several times.

[garybeck]

4. And I want to thank TIA

for the data and helping with the article

-gary

------------------------------------
the solar bus
ELECTION FRAUD AND REFORM CENTER
your home for updated information on the fight for Democracy in America
http://election.solarbus.org
------------------------------------

[davidgmills]

6. well done for the masses

In case you get any hits by Republicans, you might want to put up a link to Chuck Herrin's article and mention that some Republicans have concerns as well.

[hootinholler]

7. When you say:

There are not many possible explanations for the huge gap. One is that Mr. Bush did win the lottery and get struck by lightning on the same day — a chance in a trillion. The only other explanation is that something happened to the votes between the time they were cast by the voters and the time the official results were announced.

You are ignoring the third possibility that is being claimed by the corporatistas, the exit polls were just flat wrong. This needs to be debunked.

I suggest (edits):
There are not many possible explanations for the huge gap. One is that Mr. Bush did win the lottery and get struck by lightning on the same day — a chance in a trillion. Baring that, a second is that the exit polls are invalid. Bear with me, I'll get back to that point. The only other explanation is ...

Then further down at an apropos place point out that the polling data for State races is spot on. (I'm working from memory on that claim)
--
Then further down:

and the Republican controlled Congress who blocked legislation requiring a paper receipt for every ballot cast.

I think it's a little more forceful if you mention it was sat on in Committee for 18 Months. This is also an excellent place to point out the sudden cry for election reform by the same people sitting on these bills. Perhaps as a side-bar or footnote?
--

It is technically incorrect to say refused to let anyone see the software code that . But only technically. I'm unsure about the importance of this for your cogent article, but the point should be understood by those speaking against DREs.

In "the industry" the word 'Code' is a semi-specific pronoun. Not really a thing, but a thing-modifier (in case my horrible grammar is showing). The thing that needs to be in the public domain is the 'source code'. That is the stuff that programmers generally work on. It is then taken and run through a a series of programs which ultimately result in 'machine code'. The results of each intermediate form is also known as 'object code'.

Another important point here is that the source for the entire system should be public-domain, or at least publicly examinable. I suppose that I could be talked into accepting that tools and OSs be held private (I'm thinking Oracle or Sybase or micro$oft).

The versions of all programs actually used in elections must have valid digital signatures and public reference implementations.

With 20+ years of experience in the digital industry, I'm a paper ballot fan. I concur that Mr. Herrin's site is an excellent source.

Gary, thank you very much for distilling this! (pun intended you have a great site)

-Hoot

[garybeck]

10. thanks I updated the article

and included a brief discussion of the third possibility, citing the new rebuttal to Mitofsky's report.

thanks

[L. Coyote]

8. USCountVotes Press Release online now. New Exit Poll Report Refuted

Bear in mind that the statisticians use a very different set of numbers than those used by TIA. This report has credibility because of the numerous independent PhDs and the accurate numbers.

"
Prominent Statisticians Urge Investigation of 2004 U.S. Presidential Election Results by Kathy Dopp Monday, January 31, 2005

Prominent Statisticians Refute 'Explanation' of 2004 U.S. Exit Poll Discrepancies in New Edison/Mitofsky Report.

President Bush won November's election by 2.5% yet exit polls showed Kerry leading by 3%. Which was correct?

"There are statistical indications that a systematic, nationwide shift of 5.5% of the vote may have occurred, and that we'll never get to the bottom of this, unless we gather the data we need for mathematical analysis and open, robust scientific debate.", says Bruce O'Dell, USCountVotes' Vice President.
"

"
The statisticians go on to note that precincts with hand-counted paper
ballots showed no statistical discrepancy between the exit polls and the official results, but for other voting technologies, the overall
discrepancy was far larger than the polls’ margin of error.
"

http://www.uscountvotes.org/index.php?option=com_content&task=view&id=72&Itemid=41

[myschkin]

12. Great work!

Proposals for a second article: The west-east shift (see TIA), the machine glitches all over the country (see TIA)!

Thanks

Myschkin


PS: Maybe you could make a summary with all TIA studies (1) 2) 3) 4) 5) 6)...) - and we could send this summary to political leaders like Conyers and the media.

[Karenca]

13. Your'e right, TIA's work needs to be published

Thank you for doing this. :yourock:

[Quakerfriend]

14. Yes, thanks garybeck and TIA

for your tireless efforts!!:9

[Peace Patriot]

15. Excellent article, Gary! Very accessible and readable!

Note: Your item 1) above--misspelling of "including" (if it's fixable at this point--no biggie).


Re: the Jan. 29 report by the nine PhDs - they didn't just say that Mitofsky's theory about R's being shy of pollsters was unsupported by any evidence, they said that the exit poll data indicates the opposite of what Mitofsky said--that the exit polls actually FAVORED Bush, and they said that means that the discrepancy is probably even greater...

...concerning which, here's Freeman's analysis of predicted vote for Kerry:

-----

I'm sorry Freeman has withdrawn his 2nd paper from the internet (because of pending book publication in May), but I'm sure it's okay to quote a bit from it--it's been in circulation in draft form around here for a while. He does a brilliant calculation of the predictable Kerry vote--quite apart from the exit poll data--using the base vote from 2000, the big switch from Nader to Kerry in '04, and the new voter registration (which favored Dems by a big margin), and determines a 4 to 8 million vote discrepancy--votes Kerry should have gotten--as follows:


Table 2.2. Expected Presidential Votes based on Changes From the 2000 Election

----------Dem (G or K)---------Bush ------------3rd Party--------Tot

2000:----50,999,897 (48%)---50,456,002 (48%)---3,949,201 (4%)---105,405,100
2004:----57,890,314 (48%)---61,194,773 (51%)---1,170,071 (1%)---120,255,158

Increase:---6,890,417--------10,738,771----minus(2,779,130)-----14,850,058 (14%)


(Distributing the votes on a reasonable expectation formula:)

(1) 95%
of 00 vote----48,400,00------47,900,000---------3,800,000------100,000,000

(2) 3rd
Party -----2,300,000 (64%)------600,000 (17%)
-----------------------------------------------New voters: 20,200,000

(3) New
Voters
distrib'ed ----11,500,000 (57%)---8,300,000 (41%)

Expected
Total --------62,200,000-------56,800,000


Discre-
pancy --------(4,300,000)-------4,400.000


Freeman explains this very simply in his section entitled, "The Numbers Don’t Add Up." He says that, in 2000, Gore won the popular vote by more than half a million, but in 2004, Bush beat Kerry by 3.3 million—yet there were only two major changes in the voting population: 1) the 3rd party vote declined by 2.8 million, and 2) get-out-the-vote campaigns. 95% of the 2000 electorate voted in the 2004 election. That gives Kerry a base of 48.4 million (Gore voters), and Bush 47.9 million. Election night polls showed that Kerry got 64% of ex-Nader voters (2.5 million) and Bush got only 17% (600,000). In 2004, Dems beat Repubs in new voter registration by 57% to 41%. And when you add these three blocks of voters together—the base vote from 2000, the 3rd Party vote, and new voter registration—"…it looks as though Kerry somehow received 4,300,000 votes less than he should have, and Bush somehow received over 4,400,000 votes more than he should have."

You can read his first report, and request a copy of his second report, at:
http://www.appliedresearch.us/sf/epdiscrep.htm.

[waz_nc]

16. exit poll data

I'm not sure where to post this, but I thought some people on this board might interested. Apparently because of all the interest, they are pushing to make the raw data sets from the national exit polls available fairly soon. I've never purchased it before, but the 2002 VNS polls are available as ASCII files and as SPSS files. Codebooks and the questionnaires also available, so even if they're not available in SAS format it should be fairly easy to convert them depending if that's what you prefer.

Text of the email from the Roper Center:

We hope to have the national and state exits archived within the next week or two. We are having a special price of $99 for a CD-ROM which includes both the national and all state exits. If you want to pay be credit card, you can call me with the credit card info.

Marilyn Milliken
Roper Center
University of Connecticut
341 Mansfield Rd.
U-1164
Storrs, CT 06269-1164
[phone] Tel
[phone] Fax
Email [e-mail]
WEB www.ropercenter.uconn.edu

[RaulVB]

17. They are on schedule

They do this every time after a national election.

[RaulVB]

18. Thanks for writing this (n/t)

[Melissa G]

19. Thank you Gary and TIA!

[emlev]

20. Great article (and found a typo you might want to change)

"It is not widely know, nor is it reported"

Also, do you want to attribute the Stalin quote? It's part of its power.

Very clear article. Thank you. Don't mean to be picky, I just want it to be the best it can be!