Why the Gallup Poll numbers for Bush are much higher

[Liberatio]

My aunt sent this to me from someone she knows.  I am looking for comment on this from fellow Freepers.  Usually polls are tilted to the left.  I find this funny and sad at the same time.  To think that adults would be like this.  I wonder where this originated...

Begin email:

Please read the following and forward it to all you know, especially
those who are discouraged by the polls showing Bush ahead.  It reveals
the method used by the Gallup Poll and thereby explains why the GP
numbers for Bush are much higher than seems apparent in any other poll.
And keep registering voters -- only two weeks left for that!  Mary 

This morning we awoke to the startling news that despite a flurry of different polls this week all showing a tied race, the venerable Gallup Poll, as reported widely in the media (USA Today and CNN) today, showed George W.  Bush with a huge 55%-42% lead over John Kerry amongst likely voters.  The same Gallup Poll showed an 8-point lead for Bush amongst registered voters (52%-44%).  Before you get discouraged by these results, you should be more upset that Gallup gets major media outlets to tout these polls and present a false, disappointing account of the actual state of the race.  Why? 

Because the Gallup Poll, despite its reputation, assumes that this November 40% of those turning out to vote will be Republicans, and only 33% will be Democrat.  You read that correctly.  I asked Gallup, who have been very courteous to my requests, to send me this morning their sample breakdowns by party identification for both their likely and registered voter samples they use in these national and I suspect their state polls.  This is what I got back this morning: 

Likely Voter Sample Party IDs - Poll of September 13-15 Reflected Bush Winning by 55%-42%

Total Sample: 767

GOP: 305 (40%)

Dem: 253 (33%)

Ind: 208 (28%)

Registered Voter Sample Party IDs - Same Poll Reflected Bush Winning by 52%-44%

Total Sample: 1022

GOP: 381 (38%)

Dem: 336 (33%)

Ind: 298 (30%)

In both polls, Gallup oversamples greatly for the GOP, and undersamples for the Democrats.  Worse yet, Gallup just confirmed for me that this is the same sampling methodology they have been using this whole election season, for all their national and state polls.

 According to John Zogby himself:

 If we look at the three last Presidential elections, the spread was

34% Democrats, 34% Republicans and 33% Independents (in 1992 with Ross Perot in the race);

39% Democrats, 34% Republicans, and 27% Independents in 1996;

39% Democrats, 35% Republicans and 26% Independents in 2000.

So the Democrats have been 39% of the voting populace in both 1996 and 2000, and the GOP has not been higher than 35% in either of those elections.  Yet Gallup trumpets a poll that has consistently used a sample that shows a GOP bias of 40% amongst likely voters and 38% amongst registered voters, and depresses the Democratic portion of the sample down to levels they haven't been at since a strong three-way race in 1992?

Folks, unless Karl Rove can discourage the Democratic base into staying home in droves and gets the GOP to come out of the woodwork, there is no way in hell that these or any other Gallup Poll is to be taken seriously.

How likely is it that the Democrats will suffer a seven-point difference against the GOP this November or that the GOP will ever hit 40%?

Not very likely.

The real problem here is that Gallup is spreading a false impression of this race.  Through its 1992 partnership with two international media outlets (CNN and USA Today), Gallup is telling voters and other media by using badly-sampled polls that the GOP and its candidates are more popular than they really are.  Given that Gallup's CEO is a GOP donor, this should not be a surprise.  But it does require us to remind the media, like Susan Page of USA Today, who wrote the lead story on the poll in the morning paper, and other members of the media who cite this poll today, that it is based on a faulty sample composition of 40% GOP and 33% Democratic.

End email:

[Petronski]

Poll gurus needed here.

[Texasforever]

Yawn. We will know on Nov. 3rd.

[c-b 1]

If the Bush volunteers don't get complacent, and keep the pressure on like he is 10 points behind, the Kerry crowd is not going to know what hit them.

[Clump]

It is possible that less people are identifying themselves as dims this year. Maybe that's why random samples are getting more Republican and independent responses. That would explain the difference between polls like Zogby, Rasmussen, and Gullup, Times, and Newsweek.

[bt_dooftlook]

9/12 Democrat = Republican?

[goldstategop]

I don't think party registration matters a great deal this time. Kerry is NOT a popular figure, his campaign doesn't have a strategy or a message and Kerry is taking another vacation for a hoarse throat. Bush is well liked, has been executing a rehearsed game plan and at least is offering voters ideas for a second term and he's out their campaigning every day. The visuals tell the story. I don't need to see a poll to know who has the upper hand in the fall campaign.

[speedy]

I have thought many of the polls cited here over-sample Republicans. Either the percentage of voters calling themselves Republicans has risen dramatically in four years -- and that is a possibility -- or these polls are a set up, and later in the race will be adjusted to show Kerry closing fast on Bush to generate Dem momentum. I lean toward the second option. But I hope I am wrong, as I often am.

[Smartass]

Proud to be part of the Vast Right Wing Blogosphere  

[Drake the Lesser]

Odd. In the 1996 election, Zogby's explanation as to why his results were so accurate is that all the other pollsters over-sampled for Democrats.

[fidelio]

"Please read the following and forward it to all you know"

This is what we like to call:

[spodefly]

http://www.gallup.com/poll/content/?ci=1258

Gallup Poll Accuracy Record

Year	Candidates	Final Gallup Survey	Election Result	Gallup Deviation
		%	%	%
2000	Bush	48.0	47.9	+0.1
	Gore	46.0	48.4	-2.4
	Nader	4.0	2.7	+1.3
1996	Clinton	52.0	49.2	+2.8
	Dole	41.0	40.7	-0.3
	Perot	7.0	8.4	-1.4
1992	Clinton	49.0	43.3	+5.7
	Bush	37.0	37.7	-0.7
	Perot	14.0	19.0	-5.0
1988	Bush	56.0	53.9	+2.1
	Dukakis	44.0	46.1	-2.1
1984	Reagan	59.0	59.2	-0.2
	Mondale	41.0	40.8	+0.2
1980	Reagan	47.0	50.8	-3.8
	Carter	44.0	41.0	+3.0
	Anderson	8.0	6.6	+1.4
	Other	1.0	1.6	-0.6
1976	Carter	48.0	50.1	-2.1
	Ford	49.0	48.1	+0.9
	McCarthy	2.0	0.9	+1.1
	Other	1.0	0.9	+0.1

[flashbunny]

the fact that karl rove is mentioned in the email makes it stick out as liberal tripe.

[flashbunny]

also adding: you only need to look at the kerry campaign to know what the real poll numbers are!

[sinkspur]

Fred Barnes had an article that was posted here a couple of days ago that said, in fact, after 2002, more of the voting population was identifying itself as Republican.

Gallup is the Mercedes of polls. It's not always dead center on election day, but it's close.

Gallup is not going to destroy its reputation to throw an election to a Democrat, especially to a gnat like John Kerry.

[Eva]

They have reported a decided move to the Republican party since 9/11. Or, rather a move away from the Democrat Party. The Democrats no longer represent the average guy or gal as the case may be. There was an excellent article in the WSJ explaining this phenomon. The Democrats are now the party of the elitist academics, the big money industrialists, and the power grabbing union leaders, not to mention the gay rights and the Black power elitists. The Democrats just do not represent main stream America.

[KingKongCobra]

OK, there are two types of polls. One assumes a static electorate like Zogby, Gallup does not. If Zogby finds more republicans willing to talk on the phone, he applies his sauce to balance the poll. Gallup does not. If nothing else, the high rate of Republican response indicates a level of excitement and optimism about the race. Bush voters are energized enough to spend 1/2 hour talking to a stranger and the Kerry croud is not because they are distressed and don't really like their guy.

Which method is best? We'll see in 40 odd days but Gallup has a long history of accuracy. Note that Zogby has zero chance of determining the effect of a paradigm switch with his method. Also note that Lurch pulled funding from 4 battleground states today which have currently moved out of reach. Credible state polls (not ARG crap) indicate a major Bush shift in the last month which is consistent with Gallup.

[RWR8189]

IIRC Gallup uses a random sampling in its polls, so it would be impossible for them to oversample any group, their consistent findings of higher GOP ID over the past few months may just reflect a new reality.

Pollsters such as Rasmussen and Zogby weight the raw results they get according to party ID from exit polls in 2000. These identifications from four years ago may no longer be valid, it does not seem unreasonable to me that there has been a shift toward the GOP after 9-11.

Right when the last Gallup poll came out I reweighted the numbers to the party ID numbers that Rasmussen uses, and the results over the same time period were strikingly similar. That suggest to me that pollsters like Rasmussen and Zogby are seeing this same phenomenon where they are "over sampling" Republicans, and instead of looking any further into the issue, they just reweight the results to their own preconceived notions of what it should be.

I think that in this election pollsters that use turnout models are going to be in for a surprise. Last time Rasmussen's turnout model heavily favored Republicans and their final election day poll showed a 9 point Bush victory, this election they may miss the real shift toward the GOP and understate Bush's win dramatically.

[Brandon]

The email is false. Gallup does not now and never has weighted its polls for party I.D. Some pollsters do that -- Rasmussen, for example, assumes 39D-35R turnout, and adjusts his findings accordingly. But Gallup doesn't do that. Rather, they are finding, consistently, that more Republicans make it through their likely voter screen this year than in previous years. Someone else has already posted Gallup's track record, which is pretty darn good. Any Dem who believes the email you posted is just whistling past the graveyard.

[JohnnyZ]

It is possible that less people are identifying themselves as dims this year.

You think? After the Democrats have been opposing the troops in Iraq and elsewhere and have blown way past Howard Dean-crazy to Michael Moore Crazy? Nah, couldn't be....

One BIG thing is that many more BLACKS are self-identifying as Republicans. I don't know how much more they'll vote for Bush, but many more are confident to call themselves Republican now that they see Powell and Condi and Paige and Jackson out there as proud prominent Republicans.