Numerical Analysis: Gallup polls - May to October, 2004.

Here are the 'Rep/Dem/Ind' breakdowns and supporting internal data for the last thirteen Gallup presidential preference polls, including the October 1-3, 2004 poll which was just released. I am publishing this information in one place for easy trend comparison versus looking at each Gallup poll of Registered voters data. Also included is a re-weighting of the Gallup Registered voter polls with the year 2000 presidential election 'Rep/Dem/Ind' partisan affiliation (from VNS exit poll data) results, for comparison purposes only.

The change in the Gallup poll from the September 24-26 results of [Bush 54%, Kerry 42%, Nader 3%] to the October 1-3 poll results of [Bush 49%, Kerry 47%, Nader 1%] seem to be due primarily (say 85%) to the increase in self-identified Democrats (+4.4%), with a corresponding drop in the percentage of self-identified Republicans (-2.9%). The other factor in the Bush drop (say 15%) is due to the Independent preference going from [Bush 45%, Kerry 46%, Nader 6%] to [Bush 39%, Kerry 53%, Nader 4%]. It will be interesting to compare these Independent partisan preference internals with other polls conducted during similar time periods.

If the poll result percentages in the table show below are rounded off so as not to have a fractional component, the results should match the published polls for each polling organization. The fractional differences in the table poll results are due to the calculations done to distribute the sample number of Registered voter responses which was necessary to avoid any great inaccuracies due to the rounding already performed by the polling organizations when their results are published. Three corrections were made by me to the calculated August and September Gallup polls to correct math errors since the last time these numbers were published.



Gallup Poll Results		Republicans	Democrats	Independents	N=	Bush	Kerry	Nader	Other
All polls Registered Voters

May 7-9, 2004	100.0%	32.2%	34.1%	33.8%	877 RV	41.39%	46.41%	7.41%	4.79%
May 21-23, 2004	100.0%	34.8%	37.3%	28.0%	883 RV	44.05%	46.43%	5.55%	3.96%
June 3-6, 2004	100.0%	35.5%	37.7%	26.8%	896 RV	42.41%	45.42%	7.37%	4.80%
June 21-23, 2004	100.0%	34.1%	37.6%	28.2%	882 RV	44.90%	46.49%	5.56%	3.06%
July 8-11, 2004	100.0%	36.7%	37.4%	25.9%	891 RV	42.20%	49.61%	3.82%	4.38%
July 19-21, 2004	100.0%	40.5%	36.3%	23.1%	878 RV	43.39%	47.38%	4.90%	4.33%
July 30-August 1, 2004	100.0%	39.2%	38.3%	22.5%	916 RV	48.14%	46.94%	2.40%	2.51%
August 9-11, 2004	100.0%	38.6%	36.5%	25.0%	897 RV	45.71%	45.04%	5.13%	4.12%
August 23-25, 2004	100.0%	37.6%	33.9%	28.5%	876 RV	46.35%	46.23%	3.65%	3.77%
September 3-5, 2004	100.0%	38.2%	35.5%	26.2%	926 RV	48.49%	45.79%	3.56%	2.16%
September 13-15, 2004	100.0%	38.8%	33.0%	28.1%	935 RV	50.00%	42.00%	4.00%	4.00%
September 24-26, 2004	100.0%	39.8%	32.2%	28.0%	926 RV	52.59%	42.44%	3.13%	1.84%
October 1-3, 2004	100.0%	36.9%	36.6%	26.4%	934 RV	48.61%	47.43%	1.39%	2.57%

Gallup
Internal partisan affiliation breakdowns


							Year 2000 Presidential weighting
							Republican	Democrat	Independent

							VNS calculated exit data
		36.94%	36.62%	26.45%			34.37%	38.86%	26.77%
Poll Information	*Demographics*	Republican	Democrat	Independent		Poll Numbers	Adjusted Numbers
Gallup	Favor Bush	93.00%	10.00%	39.00%	Bush:	49.00%		46.57%
10/03/04	Favor Kerry	4.00%	88.00%	53.00%	Kerry:	47.00%		49.47%
MoE ±4.0%	Favor Nader	1.00%	1.00%	4.00%	Nader:	1.00%		1.40%
934 Registered	Other/UnDec	2.00%	1.00%	4.00%	Other/UnDec	3.00%		2.56%
		100%	100%	100%		100%		100.0%


		39.85%	32.18%	27.97%			34.37%	38.86%	26.77%
Poll Information	*Demographics*	Republican	Democrat	Independent		Poll Numbers	Adjusted Numbers
Gallup	Favor Bush	92.00%	10.00%	45.00%	Bush:	53.00%		47.64%
09/26/04	Favor Kerry	7.00%	84.00%	46.00%	Kerry:	42.00%		47.10%
MoE ±4.0%	Favor Nader	0.00%	4.00%	6.00%	Nader:	3.00%		3.35%
926 Registered	Other/UnDec	1.00%	2.00%	3.00%	Other/UnDec	2.00%		1.92%
		100%	100%	100%		100%		100.0%


		38.82%	33.05%	28.13%			34.37%	38.86%	26.77%
Poll Information	*Demographics*	Republican	Democrat	Independent		Poll Numbers	Adjusted Numbers
Gallup	Favor Bush	94.00%	11.00%	43.00%	Bush:	52.00%		48.09%
09/15/04	Favor Kerry	4.00%	85.00%	50.00%	Kerry:	44.00%		47.79%
MoE ±4.0%	Favor Nader	0.00%	0.00%	0.00%	Nader:	0.00%		0.00%
935 Registered	Other/UnDec	0.00%	0.00%	1.00%	Other/UnDec	1.00%		4.12%
	Neither	1.00%	2.00%	3.00%	Neither	1.00%		100.0%
	No Opinion	1.00%	2.00%	3.00%	No Opinion	2.00%
		100%	100%	100%		100%


		38.23%	35.53%	26.24%			34.37%	38.86%	26.77%
Poll Information	*Demographics*	Republican	Democrat	Independent		Poll Numbers	Adjusted Numbers
Gallup	Favor Bush	91.00%	7.00%	44.00%	Bush:	48.00%		45.44%
09/05/04	Favor Kerry	6.00%	90.00%	43.00%	Kerry:	46.00%		48.78%
MoE ±4.0%	Favor Nader	2.00%	2.00%	8.00%	Nader:	4.00%		3.58%
926 Registered	Other/UnDec	1.00%	1.00%	5.00%	Other/UnDec	2.00%		2.20%
		100%	100%	100%		100.0%		100.0%


		37.56%	33.90%	28.54%			34.37%	38.86%	26.77%
Poll Information	*Demographics*	Republican	Democrat	Independent		Poll Numbers	Adjusted Numbers
Gallup	Favor Bush	88.00%	8.00%	37.00%	Bush:	46.00%		43.32%
08/25/04	Favor Kerry	9.00%	88.00%	46.00%	Kerry:	46.00%		49.47%
MoE ±4.0%	Favor Nader	2.00%	2.00%	8.00%	Nader:	4.00%		3.52%
876 Registered	Other/UnDec	1.00%	2.00%	9.00%	Other/UnDec	4.00%		3.69%
		100%	100%	100%		100.0%		100.0%


		38.57%	36.45%	24.97%			34.37%	38.86%	26.77%
Poll Information	*Demographics*	Republican	Democrat	Independent		Poll Numbers	Adjusted Numbers
Gallup	Favor Bush	89.00%	6.00%	37.00%	Bush:	46.00%		42.77%
08/11/04	Favor Kerry	6.00%	87.00%	44.00%	Kerry:	45.00%		47.69%
MoE ±4.0%	Favor Nader	2.00%	5.00%	10.00%	Nader:	5.00%		5.33%
897 Registered	Do not know:	3.00%	2.00%	9.00%	Do not know:	4.00%		4.22%
		100%	100%	100%		100%		100.0%


		39.19%	38.32%	22.49%			34.37%	38.86%	26.77%
Poll Information	*Demographics*	Republican	Democrat	Independent		Poll Numbers	Adjusted Numbers
Gallup	Favor Bush	92.00%	7.00%	41.00%	Bush:	48.00%		45.52%
08/01/04	Favor Kerry	5.00%	91.00%	46.00%	Kerry:	47.00%		49.15%
MoE ±4.0%	Favor Nader	1.00%	1.00%	7.00%	Nader:	2.00%		2.66%
916 Registered	Do not know:	2.00%	1.00%	6.00%	Do not know:	3.00%		2.67%
		100%	100%	100%		100%		100.0%


		40.55%	36.33%	23.12%			34.37%	38.86%	26.77%
Poll Information	*Demographics*	Republican	Democrat	Independent		Poll Numbers	Adjusted Numbers
Gallup	Favor Bush	84.00%	7.00%	30.00%	Bush:	43.00%		39.43%
07/21/04	Favor Kerry	9.00%	89.00%	50.00%	Kerry:	47.00%		50.91%
MoE ±4.0%	Favor Nader	4.00%	3.00%	9.00%	Nader:	5.00%		5.08%
878 Registered	Do not know:	3.00%	1.00%	11.00%	Do not know:	5.00%		4.58%
		100%	100%	100%		100%		100.0%


		36.70%	37.37%	25.93%			34.37%	38.86%	26.77%
Poll Information	*Demographics*	Republican	Democrat	Independent		Poll Numbers	Adjusted Numbers
Gallup	Favor Bush	85.00%	6.00%	34.00%	Bush:	42.00%		40.60%
07/11/04	Favor Kerry	11.00%	89.00%	47.00%	Kerry:	50.00%		51.06%
MoE ±4.0%	Favor Nader	1.00%	2.00%	10.00%	Nader:	4.00%		3.91%
891 Registered	Do not know:	3.00%	3.00%	9.00%	Do not know:	4.00%		4.42%
		100%	100%	100%		100%		100.0%


		34.13%	37.64%	28.23%			34.37%	38.86%	26.77%
Poll Information	*Demographics*	Republican	Democrat	Independent		Poll Numbers	Adjusted Numbers
Gallup	Favor Bush	90.00%	8.00%	40.00%	Bush:	45.00%		44.63%
06/23/04	Favor Kerry	7.00%	86.00%	43.00%	Kerry:	46.00%		46.92%
MoE ±4.0%	Favor Nader	2.00%	3.00%	12.00%	Nader:	6.00%		5.42%
882 Registered	Other/UnDec	1.00%	3.00%	5.00%	Other/UnDec	3.00%		3.03%
		100%	100%	100%		100.0%		100.0%


		35.49%	37.72%	26.79%			34.37%	38.86%	26.77%
Poll Information	*Demographics*	Republican	Democrat	Independent		Poll Numbers	Adjusted Numbers
Gallup	Favor Bush	86.00%	9.00%	33.00%	Bush:	42.00%		41.54%
06/06/04	Favor Kerry	8.00%	82.00%	45.00%	Kerry:	45.00%		46.26%
MoE ±4.0%	Favor Nader	3.00%	5.00%	15.00%	Nader:	7.00%		7.38%
896 Registered	Other/UnDec	3.00%	4.00%	7.00%	Other/UnDec	6.00%		4.81%
		100%	100%	100%		100.0%		100.0%


		34.77%	37.26%	27.97%			34.37%	38.86%	26.77%
Poll Information	*Demographics*	Republican	Democrat	Independent		Poll Numbers	Adjusted Numbers
Gallup	Favor Bush	90.00%	9.00%	33.00%	Bush:	44.00%		43.44%
05/23/04	Favor Kerry	7.00%	83.00%	48.00%	Kerry:	46.00%		47.16%
MoE ±4.0%	Favor Nader	2.00%	4.00%	12.00%	Nader:	6.00%		5.47%
883 Registered	Other/UnDec	1.00%	4.00%	7.00%	Other/UnDec	4.00%		3.93%
		100%	100%	100%		100.0%		100.0%


		32.27%	33.98%	33.75%			34.37%	38.86%	26.77%
Poll Information	*Demographics*	Republican	Democrat	Independent		Poll Numbers	Adjusted Numbers
Gallup	Favor Bush	90.00%	6.00%	31.00%	Bush:	41.00%		41.46%
05/09/04	Favor Kerry	7.00%	87.00%	44.00%	Kerry:	46.00%		47.64%
MoE ±4.0%	Favor Nader	3.00%	4.00%	15.00%	Nader:	7.00%		6.67%
877 Registered	Other/UnDec	0.00%	3.00%	10.00%	Other/UnDec	6.00%		4.22%
		100%	100%	100%		100.0%		100.0%

Note to Cableguy: I saw your previous post on the Gallup polls, from which I have listed the data below. I could not take this data provided by Gallup to the Left Coaster website and reconcile it with the Gallup R/D/I data provided by FR poster RWR8189 in this previous post for the Gallup three-way race among Registered voters. Perhaps you could give it a try, since all the other numbers seem to fit. Best I could do was a 39.6%(R), 36.6%(D), 26.4%(I)...

Poll of October 1-3
Reflected Bush Winning by 49%-47%
934 Registered Voters

GOP: 36%
Dem: 38%
Ind: 26%

I have created a temporary 'ping' list to many who had expressed interest in these 'Rep/Dem/Ind' poll breakdowns. If you wish to be added (or removed) from this 'ping' list, please place a reply on this thread and I will attempt to clean-up this list so as to not bother those who do not wish to read this type of poll analysis.

Hope this helps...

dvwjr

Gallup poll ping...

dvwjr

Good , you weren't the only one. I tried every way possible and could not get the 36-38-36 numbers to reconcile. A 37-37-26 split looks much more likely.

Great job as usual, from my initial once-over.

SO after putting this together...what are your off-the-top-of-the-head impressions of this material? conclusions?

And which poll do you think is the most accurate of the moment? the best predictor?

That doesn't surprise me about the Left Coaster.

They have given me no reason to think they just don't make things up.

I'll tell you what concerns me. There is no question that Gallup and Newsweek heavily shifted the breakdown of voters. However, I believe that the previous breakdowns were more heavily weighted to the GOP than should have been and the current breakdown is probably more accurate. In 2000, I think that there was 4-5% more Democrats than Republicans based on exit polls. I find it hard to believe that this shifted to a 9% Republican advantage. That is a 13-14 point swing. It is more likely that we are in the same neighborhood as before, with perhaps a slight shift towards Republicans, so maybe a 2-3 point edge for Democrats, and this is exactly what Gallup & Newsweek have right now.

Regardless, since noone knows the true breakdown, we should probably be more conservative and prepare as though it is the worst case scenario.

That being said, the remarkeable shift to Kerry is bogus, due to to the changes in sampling. I think the real point to be made is that Bush never had an 11-14 point lead. It was probably more around 4-6 points (most analysts were saying this). Right now it is probably around 1-3 points. Kerry probably gained some ground.

This of course is the opinion of someone with no political analysis experience.

Gallup Ping

FReepmail me if you want to be on or off the list.

Rasmussen and Zogby show not much movement by the electorate as a whole. I think the "random" surveys are more prone to catching quick reactions to fleeting momentum, whether it be in Bush or Kerry's favor. Especially when done in conjunction with other questions, like, "who won the debate?". I think the most those polls do is drive media coverage and the emotions of far too many people.

The indies seem to favor Kerry waaaaaay too much. This seems outlierish compared to other Gallup and other recent polls. We'll have to wait and see.

Great points.

Remember some of those articles posted--I think Howard Kurtz wrote one--about the media crowd watching the debate, rooting Kerry on? They so WANTED Kerry to do well, that when Bush fumbled, they rejoiced and played it up big. For the moment, even I--and I've loathed Kerry for over 20 years--was seriously thinking "Actually, Kerry isn't THAT bad as a speaker." (Not that he'd get my vote.)

I think that's why he had that tremor in the polls--if even a huge W fan as me thought Kerry did OK, in THAT particular event, then of course people will find him acceptable.

But then the same old rhetoric comes out, and those gaffes! People can moan about how stupid W was to make faces and such, but Kerry handed Bush tons of ammo...and Bush didn't use it. NOW he's using it, and it's like the Dems wanna forget all about their great victory less than a week later.

Kerry may be a brilliant politician, but I'm starting to think W was born under a very lucky star. And a great political skill is the ability to seize and use opportunities. No one's perfect, as W demonstrated...but he sure seems to be beating Kerry on the head with the bunker-busters, Global Test, and other screwups.

Another error in my paragraphs, but not the data tables.

The best I could do was 36.9%(R), 36.6%(D), 26.4%(I).

[I had a typo of "39.6%(R)" instead of "36.9%(R)"]

dvwjr

I'm sticking with Rasmussen being the most accurate right now.

This reinforces my disbelief about Gallup.

Wasn't Rasmussen hugely off in 2000? I believe that Rasmussen predicted a Bush landslide, 49/40. I believe Rasmussen's poll is conducted by an automatic voice (as with Survey USA) as opposed to a real person.

The majority of polls (ABC/Post, NBC/WSJ, Gallup, Pew) predicted a Bush victory with a 1-3 percent margin (49/47, etc.) This margin may have been disrupted by any or all of the following:

-October surprise: Bush drunk-driving accusations
-oversampling of Republicans: may undercount minorities, people without landline telephones, poor people who work at night, etc.
-normal statistical era
-voter fraud

Take your pick of the above.

I would note, however, that Pew and ABC/Post polls suggest a 5 point lead for Bush, in contrast with CBS/NYT, Gallup, and Newsweek.

That being said, the remarkeable shift to Kerry is bogus, due to to the changes in sampling. I think the real point to be made is that Bush never had an 11-14 point lead. It was probably more around 4-6 points (most analysts were saying this). Right now it is probably around 1-3 points. Kerry probably gained some ground.

This of course is the opinion of someone with no political analysis experience.

If that is the case, why do all the other polls, Zogby,Pew,ABC/Washington Post, Rassmussien which conducted the same polls the same way WITHOUT reweighting the sample not show this Kerry Bounce"? The ones that did NOT jimmy the sample do NOT show a Kerry bounce. IF there WAS a Kerry bounce, those polls should show it.

Wasn't Rasmussen hugely off in 2000? I believe that Rasmussen predicted a Bush landslide, 49/40. I believe Rasmussen's poll is conducted by an automatic voice (as with Survey USA) as opposed to a real person.

The majority of polls (ABC/Post, NBC/WSJ, Gallup, Pew) predicted a Bush victory with a 1-3 percent margin (49/47, etc.) This margin may have been disrupted by any or all of the following:

Umm those who want to ignore 2002 and fixate on 2000 always cite Zogby as being "most accurate" and he shows a 3 point Bush lead.

I think you are getting similar numbers as the Left Coaster. Rounding errors probably explain the difference. Left Coaster claims that he gets his numbers from his contact at Gallup, and so far his numbers match well to all the previous analyses.

If Gallup's real numbers were 47.4-46.5 (rounded to 47) for Kerry and 49.4-48.5 (rounded to 49) for Bush, that would cause your numbers to fluctuate quite a bit.

Forgot to ping you. Rounding probably answers the question.

The other factor in the Bush drop (say 15%) is due to the Independent preference going from [Bush 45%, Kerry 46%, Nader 6%] to [Bush 39%, Kerry 53%, Nader 4%].

This number, through time, is volatile. You can't help but wonder what makes these independents tick, if it's post debate spin/impression then it is likely to shift as dramatically again. Bush's base and Dem support held strong, which is good news.

Kerry may be a brilliant politician, but I'm starting to think W was born under a very lucky star. And a great political skill is the ability to seize and use opportunities. No one's perfect, as W demonstrated...but he sure seems to be beating Kerry on the head with the bunker-busters, Global Test, and other screwups.

I know a majority on the Right want Bush to act like a Talk Radio host and throw 20 couterpunches for every elbow Kerry throws but did you all ever think that elections are won by more then just getting Freepers to agree with you? Bush gave Kerry the rope to hang himself with. That Global Test gaffe is bleeding him white. If the polls in Florida are any indication of reality Bush may of sealed the deal by focusing on a tour of the Hurricane damage instead of the debate. Alot of the Right never understood Clinton's appeal. They never understood how much the "Common Touch" appeals to the average Joe Sixpack vote. Bush has that common touch in spades. Kerry has it not at all. Kerry may of won on style Bush clobbered him on substance. As the Pundits get over hyperventilating at the DNC directed spin, I suspect you are going to see just how bad Kerry hurt himself Thursday night. Thursday nite Kerry demonstrated to all America what most Freepers have know all along. He is a 9-10 Candidate in a 9-11 world.

Remember some of those articles posted--I think Howard Kurtz wrote one

Howard Fineman at Newsweek. Too bad they did not listen to their own analysis in their zeal to spin a Kerry debate bounce.