Swing State Exit Poll vs. Actual Results Analysis (17/18 skewed in favor of Kerry)

Slate, Yahoo | 11/4/04 | Cableguy

[Cableguy]

			Actual				Exit Poll
	EV		Bush	Kerry	Margin		Bush	Kerry	Margin		Diff
Swing States
FL	27		52	47	5		49	51	-2		-7
PA	21		49	51	-2		46	53	-7		-5
OH	20		51	49	2		49	51	-2		-4
MI	17		48	51	-3		46	52	-6		-3
MN	10		48	51	-3		46	52	-6		-3
WI	10		49	50	-1		48	51	-3		-2
IA	7		50	49	1		49	50	-1		-2
NM	5		50	49	1		48	50	-2		-3
NV	5		50	48	2		48	49	-1		-3
NH	4		49	50	-1		44	54	-10		-9
Likely States
NJ	15		46	53	-7		44	54	-10		-3
WA	11		46	52	-6
NC	15		56	43	13		52	48	4		-9
VA	13		54	45	9
MO	11		53	46	7		52	47	5		-2
TN	11		57	42	15
AZ	10		55	44	11		55	45	10		-1
CO	9		53	46	7		50	49	1		-6
LA	9		57	42	15		56	43	13		-2
OR	7		47	52	-5
AR	6		54	45	9		54	45	9		0
WV	5		56	43	13		54	45	9		-4
ME	4		45	53	-8
HI	4		45	54	-9

 

Note: Red=Bush state/lead, Blue=Kerry state/lead

[Cableguy]

Am I the only one who sees conspiracy, liberal bias, or both?

[Cableguy]

ping

[ILS21R]

MSM kept trying until the last minute.

[atomicpossum]

It's time to junk all exit polling. It's not like getting the ACTUAL results two or three hours later is such an inconvenience (except to the media).

[Shaka]

ping.

[paudio]

I posted here before that the director of the company who did Exit Polls used to work for CBS for 2 decades.

[aft_lizard]

Looks like the bias wa about 3% average towards Kerry, so sad even there exit poll cheating couldnt help them.

[Moral Hazard]

The odds that 17/18 would be in error in one direction, and 0/18 would be in error in the other direction is so astronomically unlikely to be a result of sampling error that it's obvious there's a systemic bias in the data. Whether the source of the bias is deliberate, or reflects some sort of incompetence I don't know.

[TenaciousZ]

Am I the only one who sees conspiracy, liberal bias, or both?

By no means. But on the other hand, this very disparity could incite a new rash of stolen-election whines.

[w1andsodidwe]

Well they got Arizona right, but what are the odds they every other state would be wrong and would be wrong in Kerry's favor. Pretty unlikely.

[King Prout]

I have long held that exit polling should be banned, and that poll-results reporting should be held in abeyance until the last poll in the country had closed.

simple solution to all this mess, eh?

[Dog Gone]

We saw here clear and convincing evidence that exit polling can be manipulated for political purposes. Just as the MSM has discredited itself this year, this puts a black eye on polling organizations.

If I were an honest pollster, I'd be furious at what happened.

[Warlord]

"Am I the only one who sees conspiracy, liberal bias, or both?"

Especially when once also considers that Dem consultants and campaign workers stood ready to use the figures to create the impression of an impending Kerry sweep.

[JLS]

The odds that 17/18 would be in error in one direction, and 0/18 would be in error in the other direction is so astronomically unlikely to be a result of sampling error that it's obvious there's a systemic bias in the data. Whether the source of the bias is deliberate, or reflects some sort of incompetence I don't know.
__________________________________________________

Exactly Moral Hazard. Cableguy you really presented this data well. Showing the error always on the same side shows clearly something was wrong.

[brivette]

The efforts by the MSM to influence the election's outcome by reporting false Kerry numbers may have had the opposite effect: it made Rush, Hannity, et al, get out the base. That may have made the difference.

[aft_lizard]

I would say incompetence, but then again maybe our LAwyers and inside guys stopped massive fraud that was supposed to be covered up by these numbers.

[brivette]

What's his/her name?

[shield]

What the rats did was...send paid people to voting locations...then have them come out and look like they just voted. Then those folks gave false info about voting for kerry. This is what the rats did.

Why did they do it? To keep Republicans from going to vote (however, it fired up the Republicans even more). That didn't happen, what most likely happened was the rats heard the exit polls and they're the ones who didn't show up to vote.

The Rats cheating back fired on them big time. ;o)

[paudio]

Mitofsky.

http://www.mitofskyinternational.com/company.htm

Mitofsky International is a survey research company founded by Warren J. Mitofsky in 1993. Its primary business is conducting exit polls for major elections around the world. It does this work exclusively for news organizations. Mitofsky has directed exit polls and quick counts since 1967 for almost 3,000 electoral contests in United States, Mexico, Russia and the Philippines.

....

From 1967 to 1990, Mitofsky was executive director of the CBS News election and survey unit, and was an executive producer of its election night broadcasts. He conducted the first exit polls for CBS in 1967, and developed the projection and analysis system used successfully by CBS and Voter News Service. He started the CBS News/New York Times Poll in 1975 and directed it for CBS for its first 15 years.

[sima_yi]

If these results are correct, and if they represent all the exit polling, then I calculate (back-of-the-envelope) a probability of about 1 in 100,000 of them being due to chance alone. (17/17 wrong, all in the same direction, is about 1/2 to the 17th power.)