Numerical Analysis: Newsweek polls - March to October, 2004.

PR Newswire ^ | October 16th, 2004 | dvwjr

[dvwjr]

Here are the 'Rep/Dem/Ind' breakdowns and supporting internal data for the last eight Newsweek/PSRAI presidential preference polls, including the October 14-15, 2004 poll which was just released. I am publishing this information in one place for easy trend comparison versus looking at each Newsweek/PSRAI poll via PRnewswire links. Also included is a re-weighting of the Newsweek/PSRAI polls to the year 2000 presidential election 'Rep/Dem/Ind' partisan affiliation (from VNS exit poll data) results, for comparison purposes only.

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.

Newsweek/PSRAI  Poll  Results		Republicans	Democrats	Independents	N=	Bush	Kerry	Nader	Other
All  polls  Registered  Voters
March  18-19,  2004	100.0%	35.9%	35.1%	29.0%	838  RV	45.47%	43.44%	5.49%	5.61%
May  13-14,  2004	100.0%	32.2%	36.9%	30.9%	832  RV	42.31%	43.27%	5.41%	9.01%
July  8-9,  2004	100.0%	34.4%	35.6%	30.0%	1,001  RV	44.36%	47.35%	2.70%	5.59%
July  29-30,  2004	100.0%	29.0%	38.5%	32.5%	1,010  RV	41.78%	49.41%	3.07%	5.74%
September  2-3,  2004	100.0%	35.8%	31.5%	32.6%	1,008  RV	52.38%	40.58%	3.37%	3.67%
September  9-10,  2004	100.0%	38.0%	31.4%	30.6%	1,003  RV	49.45%	42.67%	2.19%	5.68%
September  30  -  October  2,  2004	100.0%	33.3%	36.9%	29.8%	1,013  RV	45.01%	46.59%	1.97%	6.42%
October  14-15,  2004	100.0%	36.2%	34.2%	29.7%	1,004  RV	48.31%	46.02%	1.10%	4.58%

Newsweek/PSRAI
Internal partisan affiliation breakdowns

									Year  2000  Presidential  weighting
									Republican	Democrat	Independent
									VNS  calculated  exit  data
									34.37%	38.86%	26.77%
			36.16%	34.16%	29.68%
Poll  Information		Demographics	Republican	Democrat	Independent		Poll  Numbers		Adjusted  Numbers
Newsweek		Favor  Bush	95.00%	7.00%	40.00%	Bush:	48.00%			45.78%
10/15/04		Favor  Kerry	3.00%	88.00%	51.00%	Kerry:	46.00%			48.61%
MOE  ±4.0%		Favor  Nader	0.00%	1.00%	2.00%	Nader:	1.00%			1.08%
1,004  Registered		Other/UnDec	2.00%	4.00%	7.00%	Other/UnDec	5.00%			4.53%
		0.49%	100%	100%	100%	Total:	100%			100.0%
			33.27%	36.92%	29.81%
Poll  Information		Demographics	Republican	Democrat	Independent		Poll  Numbers		Adjusted  Numbers
Newsweek		Favor  Bush	89.00%	12.00%	37.00%	Bush:	45.00%			45.08%
10/02/04		Favor  Kerry	6.00%	86.00%	42.00%	Kerry:	47.00%			47.05%
MOE  ±4.0%		Favor  Nader	2.00%	0.00%	4.00%	Nader:	2.00%			1.89%
1,013  Registered		Other/UnDec	3.00%	2.00%	17.00%	Other/UnDec	6.00%			5.97%
			100%	100%	100%	Total:	100%			100.0%
			37.99%	31.41%	30.61%
Poll  Information		Demographics	Republican	Democrat	Independent		Poll  Numbers		Adjusted  Numbers
Newsweek		Favor  Bush	93.00%	7.00%	39.00%	Bush:	49.00%			45.12%
09/10/04		Favor  Kerry	4.00%	87.00%	45.00%	Kerry:	43.00%			47.24%
MOE  ±4.0%		Favor  Nader	1.00%	0.00%	7.00%	Nader:	2.00%			1.92%
1,003  Registered		Other/UnDec	2.00%	6.00%	9.00%	Other/UnDec	6.00%			5.72%
			100%	100%	100%	Total:	100%			100.0%
			35.81%	31.55%	32.64%
Poll  Information		Demographics	Republican	Democrat	Independent		Poll  Numbers		Adjusted  Numbers
Newsweek		Favor  Bush	94.00%	14.00%	45.00%	Bush:	52.00%			49.40%
09/04/04		Favor  Kerry	4.00%	82.00%	40.00%	Kerry:	41.00%			44.12%
MOE  ±4.0%		Favor  Nader	1.00%	1.00%	9.00%	Nader:	3.00%			2.96%
1,008  Registered		Other/UnDec	1.00%	3.00%	6.00%	Other/UnDec	4.00%			3.53%
			100%	100%	100%	Total:	100%			100.0%
			29.01%	38.51%	32.48%
Poll  Information		Demographics	Republican	Democrat	Independent		Poll  Numbers		Adjusted  Numbers
Newsweek		Favor  Bush	90.00%	8.00%	39.00%	Bush:	42.00%			44.38%
07/30/04		Favor  Kerry	7.00%	86.00%	45.00%	Kerry:	49.00%			47.53%
MOE  4.0%		Favor  Nader	1.00%	1.00%	7.00%	Nader:	3.00%			2.76%
1,010  Registered		Other/UnDec	2.00%	5.00%	9.00%	Other/UnDec	6.00%			5.33%
			100%	100%	100%	Total:	100%			100.0%
			34.37%	35.66%	29.97%
Poll  Information		Demographics	Republican	Democrat	Independent		Poll  Numbers		Adjusted  Numbers
Newsweek		Favor  Bush	90.00%	10.00%	34.00%	Bush:	44.00%			43.59%
07/09/04		Favor  Kerry	6.00%	83.00%	53.00%	Kerry:	47.00%			48.32%
MOE  ±4.0%		Favor  Nader	1.00%	1.00%	6.00%	Nader:	3.00%			2.54%
1,001  Registered		Do  not  know:	3.00%	6.00%	7.00%	Do  not  know:	6.00%			5.56%
			100%	100%	100%	Total:	100%			100.0%
			32.21%	36.90%	30.89%
Poll  Information		Demographics	Republican	Democrat	Independent		Poll  Numbers		Adjusted  Numbers
Newsweek		Favor  Bush	85.00%	12.00%	35.00%	Bush:	42.00%			42.94%
05/14/04		Favor  Kerry	6.00%	78.00%	42.00%	Kerry:	43.00%			43.20%
MOE  ±4.0%		Favor  Nader	3.00%	2.00%	11.00%	Nader:	5.00%			5.06%
832  Registered		Other/UnDec	6.00%	8.00%	12.00%	Other/UnDec	10.00%			8.80%
			100%	100%	100%	Total:	100%			100.0%
			35.92%	35.08%	29.00%
Poll  Information		Demographics	Republican	Democrat	Independent		Poll  Numbers		Adjusted  Numbers
Newsweek		Favor  Bush	88.00%	11.00%	35.00%	Bush:	45.00%			43.74%
03/19/04		Favor  Kerry	8.00%	81.00%	43.00%	Kerry:	43.00%			45.41%
MOE  ±4.0%		Favor  Nader	1.00%	4.00%	12.00%	Nader:	5.00%			5.35%
838  Registered		Do  not  know:	3.00%	4.00%	10.00%	Do  not  know:	7.00%			5.50%
			100%	100%	100%	Total:	100%			100.0%

Source: Newsweek poll, October 14-15, 2004

Source: Newsweek poll, September 30 - October 2, 2004

Source: Newsweek poll, September 9-10, 2004

Source: Newsweek poll, September 2-3, 2004

Source: Newsweek poll, July 29-30, 2004

Source: Newsweek poll, July 8-9, 2004

Source: Newsweek poll, May 13-14, 2004

Source: Newsweek poll, March 18-19, 2004

It will be interesting to compare these internals with other polls conducted in similar time periods. Look below at the Newsweek poll of October 14-15, 2005 of 1,004 'Registered' voters as compared to the Washington Post tracking poll of October 12-14, 2004 of 1,203 'Likely' voters. Quite a swing between the Newsweek registered and Washington Post likely voter internals...

Poll  Information		Demographics	Republican	Democrat	Independent		Poll  Numbers
Newsweek		Favor  Bush	95.00%	7.00%	40.00%	Bush:	48.00%
10/15/04		Favor  Kerry	3.00%	88.00%	51.00%	Kerry:	46.00%
MOE  ±4.0%		Favor  Nader	0.00%	1.00%	2.00%	Nader:	1.00%
1,004  Registered		Other/UnDec	2.00%	4.00%	7.00%	Other/UnDec	5.00%
			100%	100%	100%	Total:	100%

Poll  Information		Demographics	Republican	Democrat	Independent		Poll  Numbers
WaPost  Tracking		Favor  Bush	90.00%	12.00%	46.00%	Bush:	48.00%
10/14/04		Favor  Kerry	10.00%	85.00%	46.00%	Kerry:	48.00%
MOE  ±3.0%		Favor  Nader	0.00%	1.00%	5.00%	Nader:	1.00%
1,203  Likely		Other	0.00%	0.00%	1.00%	Other:	0.00%
		Neither	0.00%	0.00%	2.00%	Neither:	1.00%
		Wont  Vote	0.00%	0.00%	0.00%	Wont  Vote:	0.00%
		Do  not  know	0.00%	2.00%	2.00%	Do  not  know:	1.00%
			100%	100%	102%		99%

Source: Washington Post - October 12-14, 2004

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

[dvwjr]

Newsweek poll ping...


dvwjr

[NormsRevenge]

Democratic presidential nominee John Kerry (news - web sites) holds up the pumpkin he chose at the Garringer Family Pumpkin Patch in Jeffersonville, Ohio October 16, 2004, part of a day-long campaign bus trip through Ohio. REUTERS/Brian Snyder

[Owen]

Good work, but . . . realize it may all be meaningless. The Dem/Rep/Ind question asked may not be "Are you registered Dem/Rep/Ind". It may be "Do you best identify with Dem/Rep/Ind", in which case the answer could very well change every week or two.

If so, partisan mix on each poll becomes a lot less important a thing to notice.

[Darkwolf377]

Great stuff, thanks again.

If I'm reading this right it shows that Kerry's never reached the highs in the poll that Bush has, at any point (i.e. never broke 50%).

[GSlob]

Now... which one is the pumpkin?

[Right_in_Virginia]

Are "registered voters" as good a barometer as "likely voters"?

[MNJohnnie]

If so, partisan mix on each poll becomes a lot less important a thing to notice.


I have explained this to you three times. YES the partisan mix DOES matter. If you accept MORE responses from people who identify themselves as Republicans, as this poll does, you get a stronger pro Bush showing. IF you inteview more DEMOCRATS, you get a worse showing for Bush. WARNING, Newsweek is AT it again. They OVERsampled Republicans this time to show a Bush Lead. NOW in a week to 10 days, they will REVERSE the sample and scream for all their are worth about "Bush's fading poll numbers" More DNC media spin being set up courtesy of Newweak.

[MNJohnnie]

WARNING, Newsweek is AT it again. They OVERsampled Republicans this time to show a Bush Lead. NOW in a week to 10 days, they will REVERSE the sample and scream for all their are worth about "Bush's fading poll numbers" More DNC media spin being set up courtesy of Newweak.

[Right_in_Virginia]

I bet if we did a search of the pictures taken by Brian Synder during this campaign we'd owe him a beer ;-)

[Cableguy]

There is something seriously wrong with this Newsweek poll. For example, Bush has a clear advantage with women, who prefer him 49 percent to 43 percent. Kerry has a slight edge with men, 50 percent to 46 percent.

Newsweek is widely distrusted by almost everyone in the polling world. You can now see why no one takes their results seriously. Last Newsweek poll was flawed, and so is this one.

[dc-zoo]

Its only 2%.. 36.2% Rep and 34.2% Dems. Doesn't appear as such a big deal to me.

[Numbers Guy]

There is something seriously wrong with this Newsweek poll. For example, Bush has a clear advantage with women, who prefer him 49 percent to 43 percent. Kerry has a slight edge with men, 50 percent to 46 percent.

This poll is almost laughable, in spite of the reasonable overall numbers.

[dvwjr]

Yep. Not only is the male/female ratio for Bush way off from previous polls, but I don't believe that their sample of Independents favors Kerry over Bush by 11% points. I cannot find the internals of any other poll within the last three or four weeks which shows such an advantage.

Not a good poll, but I'll continue to dissect the R/D/I numbers because they will trend towards the 'Dems' until the final pre-election poll that they publish...


dvwjr

[MNJohnnie]

Its only 2%.. 36.2% Rep and 34.2% Dems. Doesn't appear as such a big deal to me.

Yes, it is when you have a 1 point race among registered voters. Flip those numbers and suddenly we go into election week with "Kerry surging to a 4 point lead" in the Newsweak poll. Remember 2% is an overnight 4% swing. Bush down 2 Kerry up 2. And that is assuming they don't do what they did before and suddenly kick it up to 39% Dims. But I have to go to bed now. Been on line too long and am starting to take these polls too serious. IT IS ALL ABOUT turnout. If Dims want to play games with the POLLS let them. The Republican base has seen the REAL face John Kerry is hiding under his "moderate" mask and we will be walking barefoot across ground glass election day to relect GW Bush.

[Future Useless Eater]

dang, I smell a photoshop coming...

[NormsRevenge]

I rack 'em..

You smack em. ;-)

[ride the whirlwind]

Thanks for the dissection.

[Owen]

Guy, relax. They are not Democrats. They are not Republicans. They are asked a question "In today's politics, with which party do you most identify?"

The answer to that question can widely vary from week to week because it's a question about sentiment. This means that the 2000 VNS Exit poll data that is the target partisan mix for most people trying to define the breakdown of party numbers in the US is not valid to compare to these polls because that question was "Are you a registered Democrat/Rep/Ind?"

These are two totally different questions. The wide variance we are seeing in partisan mix is not contrived by any of the pollsters. They make the random calls and they ask the first question above, and when Bush has a good day that day there is a shift of sentiment in the conservative direction and presto, the sample results in more "GOP" respondents. The quote marks around GOP are the difference.
The ratio of *registered* Dems to *registered* GOP may not be varying anything at all like what the polls report.

If next week there are more "Dems" appearing in the samples and Kerry gets a surge, it won't be because it was contrived by the pollster. It will be because sentiment shifted.

We have to be careful in how we describe this. There are not more Dems or GOP or Inds in these samples. There are more people who identify their ideals on that particular day with those parties in the samples. This is even more complicated because some polls might very well be asking question number 2 and not number 1. But note that this is difficult because some states have no partisan registration.

[AndyTheBear]

Kerry hasn't broken 50 in just about any poll I've seen on realclearpolitics.com. By contrast Bush has done it dozens of times.