Bush Exit Poll Job Approval: A smoking gun?

[TruthIsAll]

The average of 11 final Pre-election polls had Bush approval
at 48.5%.

In the preliminary Exit Poll (11027 of 13660 final
respondents) the Bush Approval question was answered by n=5666
respondents, who gave Bush their own 51% approval rating.

Using this job approval as a demographic weight, Kerry won by
almost 3.6 MILLION VOTES.

Bush Approval
	 Mix   Bush	Kerry	Nader
Approve	51%	88%	11%	1%
Dissap	48%	5%	93%	1%

Total	  99%	47.28%	50.25%	0.99%
Votes	  120.5  57.8 61.44	1.21

The MoE for 5666 respondents = 1.33%. 
............................................................

Assuming the Nov. 1 Bush 48.5% job approval was closer to the
truth, Kerry won by almost 10 MILLION VOTES.

Bush Approval
	  Mix  Bush	Kerry	Nader
Approve	48.5%	88%	11%	1%
Dissap	51.5%	5%	93%	2%

Total	  100%	45.26%	53.23%	1.52%
Votes 122.2   55.33 65.08 1.85

[TruthIsAll]

1. Here is a link to the 11027 National Exit Poll...

http://www.exitpollz.org/mitof4zone/US2004G_3798_PRES04_NONE_H_Data.pdf

[LatePeriduct]

2. Unbelievable actually....

What would be really telling, is if you could match the samples and find out where those 10 million votes were coming from.

See if you could cross-check that with the final poll where 10 million dissapeared into the Bush column, that might narrow down the exact city districts.

[Old and In the Way]

3. First, I don't believe 9/10 Republicans voted for Bush.

My own little informal poll of known Republicans in rural central Maine right after the election, I found that, of 10 registered Republicans, only 2 voted for George Bush. Now, I'm sure that this was somewhat of an aberation....but the point here (and I remember lots of stories from other DUers about Kerry voting Repuublicans) is that I find 88/100 Republicans would not vote for Bush over Kerry...no way, no how.

But I'm convinced if you went into the polling booth and were voting electronically, good chance you voted Bush if you were registered Republican.

This would be an interesting analysis, TIA: controlling for voting system type, what were the % of Bush voters. I know the answer. Paper ballots would show a lot less Bush voters than electronic ballots. Still, it would be good to show the correlation by state, by type....

[bitchkitty]

10. Not an aberation -

you are correct. Think of the military vote, and why we haven't seen the results anywhere - I think they went for Kerry, big time.

[On the Road]

15. Yes, That Would a Great Study

Increasing the proportion of Bush votes among Republicans would be a low-visibilty way of increasing the vote.

This is also why paper receipts are not enough -- the paper has to be able to be tied to the record of and individual vote by an ID field of some kind. Otherwise you can't do many of the most important validations.

[Old and In the Way]

18. What we really need is..............

Have Howard Dean and the DNC do a "Who Did You Really Vote For in "04"
Audit....Big time....make it a very big deal. Require everyone who wishes to "vote again" to sign an affidavit. Make it really, really worth something.

America's Revotes This time, we control the terms of the vote. I'd do it....in a heartbeat. Let's see if the results are the same. Let's compare a Democraticly-controlled vote vs. Republican controlled voting. You know what? I think we are into honesty and if Bush wins....so be it. Let's find out. Republicans are, of course, welcome to participate.

I'm a "vote again"-typr of Democrat.

[Kurovski]

4. Once again, thank you TruthIsAll.

Recommended for greatest page.

[fooj]

5. Thank you for never, ever giving up! NGU!!!

Peace.:patriot:

[Nothing Without Hope]

6. Here are some other intriguing breakouts from the same exit poll

these are all from TIA's data. I'm excited about them because I finally saw the obvious patterms for myself: It's so clear that not only did Kerry win, he won exactly the way we expected him to.

In each group of five numbers, the first number is the PERCENTAGE OF THE TOTAL VOTES ANALYZED, the second is % votes for KERRY, the third is % votes for BUSH, and the fourth is % votes for NADER. (TIA thinks the fifth number (“Rep2000”) is probably the difference in % Bush votes compared to 2000)

Total votes: 51% Kerry, 48% Bush, 1% Nader.
Females were pro-Kerry (9% margin) while males were more narrowly pro-Bush (4% margin)

Number of Interviews - 11,027
Vote Estimate Not for On-Air Use
100 51 48 1 -
Are you: (n=10,978)
Male 46 47 51 1 -2
Female 54 54 45 1 2

Kerry won STRONGLY among first-time voters (13% margin), who represented 11% of the total votes. In this analysis, Kerry also won among the respondents who had voted before (2% margin)

Is this the first time you have ever voted? (n=5,430)
Yes 11 56 43 - 0
No 89 50 48 1 0

Self-described independents made up 26% of total and went for Kerry (7% margin). There were slightly more Dems than Repubs (38% vs 36%) and both went strongly for their party’s candidate.

No matter how you voted today, do you usually think of yourself as a:(n=10,553)
Democrat 38 90 9 1 -2
Republican 36 7 92 0 1
Independent or something else 26 52 45 2 -2

People deciding on their choice in the last month (15% of total) went more for Kerry. People who decided before that (79%) were evenly divided, 50/50.

When did you finally decide for whom to vote in the presidential election? (n=6,856)
Just today 6 53 40 5 -
In the last three days 3 53 41 4 -
Sometime last week 2 48 51 1 -
During the last month 10 60 38 1 -
Before that 79 50 50 0 –

Here is the breakdown by four sections of the country. Kerry won by 18% margin in the East and by an 8% margin in theWest; Bush won by a 1% margin in the Midwest, and Bush won by a 9% margin in the South.

National Region (n=11,027)
East 22 58 40 1 1
Midwest 25 50 49 1 0
South 31 45 54 1 -1
West 21 53 45 1 -1

Kerry won by a 19% margins in larger cities (pop. over 50,000) and by a 1% margin in suburbs; Bush won by an 8% margin in small cities and rural populations.

Population of area, three categories (n=9,027)
City over 50 thousand 30 59 40 1 5
Suburbs 45 50 49 1 0
Small city and Rural 24 45 53 1 -6

[EFerrari]

7. Thank you, TIA

for making me feel less crazy all these months.

I came to DU when nothing matched what I felt happened in this election.

Now that I know, beyond a reasonable doubt, that it was stolen, I know where to put my energy.

Thanks for your work all this long time.

B.

[bushisanidiot]

8. Repukes can't win without cheating. AWOL Bush drove away thousands

of repukes from the polls... many didn't even show up because they couldn't vote for either candidate. Yet, hundreds of thousands newly registered democrats pushed voter registration numbers through the roof!! no way in hell AWOL Bush won with his 48% approval rating.

[On Par]

9. I'd Love To Get It, But I Don't !

I don't know what "mix" is. While you're showing me 88%/5% and 93%/11% in both polls with different poll ratings for Bush, I don't understand how you made the computation for 3.5 and 10 million. Also, for me, n='s was always the unknown number.

I did far better with your Tragedy in 3 Acts. For this deliniation, I need a ledger. I need to see how the computations were derived. Thus far, I can only assume that the difference in the Bush polls were somehow calculated to get the margin for Kerry.

I want to believe. I just can't figure it out. In other words, you'd never find me saying "Alex, I'll take "exit polls" for $200.

[TruthIsAll]

11. Look at the last line- 122.2 million votes. 53.23% of 122.2 = 65.08 mm

Assuming the Nov. 1 Bush 48.5% job approval was closer to the
truth, Kerry won by almost 10 MILLION VOTES.

Bush Approval
	  Mix  Bush	Kerry	Nader
Approve	48.5%	88%	11%	1%
Dissap	51.5%	5%	93%	2%

Total	  100%	45.26%	53.23%	1.52%
Votes    122.2  55.33   65.08   1.85

[On Par]

12. I Need A Course In Exit Polling...

I just looked at your probabilites for all 14 states going to Bush and the MOE, and understood it completely.

Okay, I see how you got this computation and 10 million.

Just one more stupid question. What's "Mix?"

[TruthIsAll]

13. Mix is just a term for the demographic breakdown. Total mix = 100%

2004 voters

Gender mix:

Females = 54%
Males = 46%

54% of 122.26 million =66.02
46% of 122.26 = 56.24

[On Par]

17. Thanks. You've Been A Fine Teacher to this untrained...

...math interpreter.

[Peace Patriot]

14. I'm struggling with this, too. Your conclusion doesn't surprise..

...me in the least. It is consistent with everything else we know (the Zogby prediction, the current and past opinion polls (Bush approval at 49% on his Inauguration Day!, and 45% last week, with numbers in the 60% to 70% range disapproving all major Bush policy, and consistent unprecedented disapproval over time, in many different polls). But I don't quite follow your argument here.

Let me break down my questions, item by item:

"The average of 11 final Pre-election polls had Bush approval
at 48.5%."

You're speaking here of opinion polls from different polling organizations, prior to the election, which you have averaged together? Right? Just to be clear. Do you have a date for these?

"In the preliminary Exit Poll (11027 of 13660 final
respondents) the Bush Approval question was answered by
n=5666 respondents, who gave Bush their own 51% approval
rating."

The "preliminary Exit Poll" is early on election day? Of actual voters, early in the day (right)? I'm kind of confused, in that case, why they were asked for a Bush approval rating, rather than asked how they just voted? (Or maybe they were asked both? --and if so, were all voters who were exit polled asked for both a Bush approval rating, and how they voted?). If they were asked both, was the 51% approval rating in this group consistent with their vote? Are these preliminary Exit Poll respondents a subgroup of the "final respondents"?

"...who gave their own 51% approval rating." What does "their own" mean? The Exit Pollsters decided to treat this early group in a special way (i.e., asking for approval/disapproval of Bush, as compared to....?)?

I guess what I'm asking here is: WHAT is a "preliminary Exit Poll"? (I'm not sure.) Are these actual voters, and, if not, how were these respondents selected? Was it just an opinion poll by another name? (And, when were they polled?)

------

"Using this job approval as a demographic weight, Kerry won by
almost 3.6 MILLION VOTES."

What do you mean by "this" ("Using THIS job approval...")? (The average of the preelection opinion polls? The "preliminary Exit Polls"? A combination?)

How are you getting from the percentages (i.e., 51% Bush approval in prelim Exit Polls) to the Kerry vote (won by 3.6 million)?

Where does the 48% "Dissap" figure, in the "Mix" column, come from? (And, what does "Mix" mean? I really don't understand it.)

Bush Approval
Mix Bush Kerry Nader
Approve 51% 88% 11% 1%
Dissap 48% 5% 93% 1%

Total 99% 47.28% 50.25% 0.99%
Votes 120.5 57.8 61.44 1.21

The MoE for 5666 respondents = 1.33%.
............................................................

"Assuming the Nov. 1 Bush 48.5% job approval was closer to the
truth, Kerry won by almost 10 MILLION VOTES."

You are talking now, in saying "Nov. 1," about the "average of 11 final Pre-election polls"? Just to be clear. All 11 were conducted on the day before the election? Given this, what then was the "preliminary Exit Poll," and when was IT conducted? (--I'm having to assume that "closer to the truth" means "closer to the election," i.e. Nov. 1; ergo the "preliminary Exit Poll" must have been conducted prior to Nov. 1. But if that's true, why is it called an "Exit Poll"?)

Bush Approval
Mix Bush Kerry Nader
Approve 48.5% 88% 11% 1%
Dissap 51.5% 5% 93% 2%

Total 100% 45.26% 53.23% 1.52%
Votes 122.2 55.33 65.08 1.85

Again, I guess I'm just not sure of the terms of this table, and where each figure came from. Approve? Dissap? Mix? Also, didn't Kerry WIN in the early Exit Polls--the data that appeared on everybody's TV screens (that was fiddled later in the day to fit the official tally)? Why is there a discrepancy between Bush approval (51%) in the "preliminary Exit Polls" and the early on election day Exit Polls (of actual voters)? (And, to repeat my question from above: WHO was being polled in the "preliminary Exit Polls," and when?)

[TruthIsAll]

16. Lots of questions.

"The average of 11 final Pre-election polls had Bush approval
at 48.5%."

You're speaking here of opinion polls from different polling organizations, prior to the election, which you have averaged together? Right? Just to be clear. Do you have a date for these?

Around Nov.1. These are the final ratings:

Rating Avg Nwk Fox Gallup Pew Harris CBS ABC Time NBC AP Zogby
Current 48.5 46.0 49.0 51.0 44.0 na 44.0 53.0 53.0 49.0 47.0 49.0


"In the preliminary Exit Poll (11027 of 13660 final
respondents) the Bush Approval question was answered by
n=5666 respondents, who gave Bush their own 51% approval
rating."

The "preliminary Exit Poll" is early on election day? Of actual voters, early in the day (right)? I'm kind of confused, in that case, why they were asked for a Bush approval rating, rather than asked how they just voted? (Or maybe they were asked both? --and if so, were all voters who were exit polled asked for both a Bush approval rating, and how they voted?).

It was just one of many questions asked: Do you approve or disapprove of Bush? It was NOT meant to be a rating poll per se. The National exit poll is really a set of individual polls.


If they were asked both, was the 51% approval rating in this group consistent with their vote?
yes

Are these preliminary Exit Poll respondents a subgroup of the "final respondents"?
Yes


"...who gave their own 51% approval rating." What does "their own" mean?
It was their rating because they gave it.

The Exit Pollsters decided to treat this early group in a special way (i.e., asking for approval/disapproval of Bush, as compared to....?)?

You are reading too deeply into this. It's just another measure of voter characteristics.

I guess what I'm asking here is: WHAT is a "preliminary Exit Poll"? (I'm not sure.) Are these actual voters, and, if not, how were these respondents selected? Was it just an opinion poll by another name? (And, when were they polled?)

The term preliminary refers to the fact that it was a completed subset of the final? But it is all part of ONE poll:

1. 8349 respondents
2. 11027
3. 13047
The poll should have ended here, at 13047.

The final 613 (the 13660) were an artifice, and not to be believed - a ploy to match a bogus vote. But this the "official" poll. This is what the media wants you to accept.



------

"Using this job approval as a demographic weight, Kerry won by
almost 3.6 MILLION VOTES."

What do you mean by "this" ("Using THIS job approval...")? (The average of the preelection opinion polls? The "preliminary Exit Polls"? A combination?)

I use 51% Bush approval based on the first 11027 to calculate the weighted Kerry vote - it is consistent with all the other demographics in the 11027 summary - 51-48%, or 3 million votes.

Then I use the 48.5% average as a weight, and Kerry wins by 53-45%, or 10 million votes.

How are you getting from the percentages (i.e., 51% Bush approval in prelim Exit Polls) to the Kerry vote (won by 3.6 million)?

Multiply the 50.25% by the 122.26 million votes= 61.43 million.

Where does the 48% "Dissap" figure, in the "Mix" column, come from? (And, what does "Mix" mean? I really don't understand it.)

48% disapprove of Bush
Mix is my term for the demographic weighting. The total mix should equal 100%, but it's 99%, probably due to roundoff error.

Bush Approval
Mix Bush Kerry Nader
Approve 51% 88% 11% 1%
Dissap 48% 5% 93% 1%

Total 99% 47.28% 50.25% 0.99%
Votes 120.5 57.8 61.44 1.21

The MoE for 5666 respondents = 1.33%.
............................................................

"Assuming the Nov. 1 Bush 48.5% job approval was closer to the
truth, Kerry won by almost 10 MILLION VOTES."

You are talking now, in saying "Nov. 1," about the "average of 11 final Pre-election polls"? Just to be clear. All 11 were conducted on the day before the election? Given this, what then was the "preliminary Exit Poll," and when was IT conducted? (--I'm having to assume that "closer to the truth" means "closer to the election," i.e. Nov. 1; ergo the "preliminary Exit Poll" must have been conducted prior to Nov. 1. But if that's true, why is it called an "Exit Poll"?)

Again, the term preliminary is a misnomer. It is the first 11027 respondents of the TOTAL EXIT POLL taken on Election Day, Nov.2.

I say closer to the truth because the 48.5% was an average of 11 separate pre-election polls, as shown above. Assuming the average sample-size was 1000, then 10,000 were polled

Bush Approval
Mix Bush Kerry Nader
Approve 48.5% 88% 11% 1%
Dissap 51.5% 5% 93% 2%

Total 100% 45.26% 53.23% 1.52%
Votes 122.2 55.33 65.08 1.85

Again, I guess I'm just not sure of the terms of this table, and where each figure came from. Approve? Dissap? Mix? Also, didn't Kerry WIN in the early Exit Polls--the data that appeared on everybody's TV screens (that was fiddled later in the day to fit the official tally)? Why is there a discrepancy between Bush approval (51%) in the "preliminary Exit Polls" and the early on election day Exit Polls (of actual voters)? (And, to repeat my question from above: WHO was being polled in the "preliminary Exit Polls," and when?)

LETS BE CLEAR: 48.5% IS THE AVERAGE OF 10 FINAL PRE-ELECTION POLLS.
51% IS THE RESULT FROM THE EXIT POLL TAKEN ON THE DAY OF THE ELECTION AFTER THE RESPONDENTS VOTED.

I guess I am assuming a certain level of familiarity with the terminology, pre-election Bush ratings and Exit Poll categories on the part of the general DUer who has followed my posts for months. now.

[TruthIsAll]

19. Usually, when someone does a lousy job, they get fired,,

.