Posted on 09/21/2004 5:20:54 PM PDT by dvwjr
Here are the latest political affiliation breakdowns for many of the major polls that have been published. This allows one to judge how much any variance in the political affiliation of the pool of poll respondents affects the final published poll results. It will allow for a comparison to those future polls which will be released in the next 42 days. Updated figures are for the ABC/Washington Post, American Research Group, CBS News, Fox News/Opinion Dynamics, Gallup, Los Angeles Times, Newsweek/PSRAI and Time/SRBI sponsored polls of Registered voters for a three-way race (Bush vs Kerry vs Nader) in the presidential contest.
(Excerpt) Read more at washingtonpost.com ...
Please add me to your list.
Zogby mixes his analysis of other valid state polls with his 'interactive' national poll which is worthless.
Currently Bush has a 5%-6% point lead over Kerry and Nader in the national 'popular' vote projections. While it is true that the State by State electoral votes are the only ones which count, the "popular" vote as a quick and dirty yard-stick is still important.
Dick Morris, despite his abysmal prediction and analsys record IS very good with numbers. In the 2000 election he said that a model he did for Clinton in 1996 determined that the possiblility of a electoral college 'win' and popular vote 'loss' could only take place if the national popular vote was 0.5%, the election was seperated by the 0.51% final Bush/Gore result.
For the year 2004 model, Morris has changed to a 0.70% model for a split decision election. It will be interesting to see if his new numbers for his model hold up. Of course, only in another very close race would his theory be put to the test.
It's all about vote distribution:
In 1976 Jimmy Carter got 50.08% of the 'popular' vote and 55.20% of the Electoral College vote.
In 1980 Ronald Reagan got 50.75% of the 'popular' vote and 90.89% of the Electoral College vote.
Reagan's extra 0.67% of the 'popular' vote got him an additional 35.69% of the Electoral College vote...
dvwjr
That's not necessarily a problem -- these are self-identified. After a convention it's not uncommon for some of the public to develop a bit of "me-too"-ism and claim that particular party affiliation. I don't think party affliation should be weighted, especially for the pool of likely voters, but the numbers do need to be taken into some consideration. Remember, momentum is the key.
I don't like the movement of the Newsweek poll, but compared to the others it looks like it may be an outlier.
The numbers I just pulled out of my behind show
Bush 49.8
Kerry 44.1
anyone else noticed the habitual over sampling of dems and undersampling of reps - especially by aBS/washsocialist?
I thought Gallup didn't weigh party affiliation.
34.37% Rep 38.86% Dem 26.77% Ind
then the results would be:
45.43% Bush 47.20% Kerry 01.97% Nader 05.41% Other
For what it's worth...
dvwjr
Hope this helps...
dvwjr
thanks for the ping.
Thanks, interesting analysis. Keep me on your ping list.
Great stuff! Please keep me on your ping list.
please put me on yor ping list, good work here
great charts - bump for reference
>>I take it you were re-weighting the latest Newsweek poll dated September 9-10, 2004. If I use the VNS calculated year 2000 exit poll data of the R/D/I splits...
Don't forget that there has been a LARGE swing in R/D/I 'party ID' (party Identification)
towards the Republicans since the 2000 election.
Nationwide AVERAGES show that Republicans have picked up a 5 point advantage since the 2000 election:
http://www.realclearpolitics.com/Commentary/com-11_11_03_MK.html
...OTHER outstanding POLL commentary here - NOT to be missed:
http://www.realclearpolitics.com/commentary.html
Agreed. I don't think that the year 2000 R/D/I splits will be confirmed with the exit poll data from the 2004 election.
As to methodology, who is correct? The 'Adjusters' or the 'Leave-it-Alones'???
The 'Adjusters' would have never caught the huge shift of the voting public from Republican to Democrat during the 1932 election campaign in both the Congress and the Presidency (except for the obvious fact of the Great Depression), whereas the 'Leave-it-Alones' numbers would have seen the tidal wave coming.
It might be explained by the fact that the long awaited shift of the public behind the Republicans as the MAJORITY party may be underway.
Or not.
dvwjr
That, is the million dollar question. Should their be weighted adjustments or just simple polls.
Zogby weights too much (and note that he is not using the 2002 weights).. then again, maybe he is right, maybe we are behind.
I find that very, very hard to believe based on all the other polls and what I see both in the strategy being used by the candidates and their rhetoric.
Maybe just a Virginian... ;-)
dvwjr
would like to be on the ping list too...
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.