Gallup Final Estimate: O 55 - M 44

Gallup ^ | 11/02/08 | Gallup

[ubaldus]

Obama +11 with undecideds allocated. Without allocation: RV O 53 - M 40, LV O 53 - M 42 (in both models).

[tatown]

History is against you if you think dem turnout will be anything greater than 4% points.

The most accurate pollster of the 2004 presidential election is assuming D +3. Maybe you can explain to us why they are wrong especially when the dem blowout year of 2006 had Dems at +3?

[SallyH]

“Columbus OH news: Palin in Ohio today, a crowd of 3500 Obama in Ohio today, a crowd of 65,000”

Palin was in an airport hanger where attendance was limited to 3500-—that’s all the tickets they gave out. She was doing like 5 quick rallies today.

Bruce Springsteen was with Obama....no contest...Bruce wins there.

[latina4dubya]

I think the true gap is 5 to 8 points in this election.

no way... they're not showing any of that in early voting numbers, and early voting was a big Obama/Democrat to-do... it's not happening... no youth vote... nada, enchilada...

[bailmeout]

I think it is legit to assume a 3-5% spread in favor of the Dems.

Every action has an equal and opposite reaction. IMO, ACORN + large numbers of pro-Dem early voters + media effort to coronate BHO before the election = major increase in GOP and pro-McCain Dem turnout on election day. I wouldn't be shocked if turnout is equal or slightly pro-GOP.

[rwilliam]

I think D +10 party ID gap is probably an overestimate. Rasmussen shows D +6.5 in his 6-week average, I think the true gap is 5 to 8 points in this election.

What leads you to believe that the gap in party affiliation will go from +2 for the Democrats in 2006 to +5 or +8 in 2008? Do you really believe that the Democrats are that much more motivated to vote on Tuesday compared to Republicans?

Also, Rasmussen released data yesterday to suggest that Democrats are much more willing to take part in exit polls. However, he must be assuming that the same trend doesn't exist for standard polling. If that assumption is wrong, his assumption about party affiliation (which is then enforced in his polls through weighting) is wrong.

-Bob

[Greg67]

I came in this country for The First Amendment

NO BANS for any reason ...free market works in poling too ...

And yeah for The Second two.

[StarFan]

Thanks for clearing that up.

[kesg]

[beagleone]

Gallup can go gallup their kool-aid asses right on outta here.

[kesg]

Gallup has swallowed the Koolaid and is now asking for seconds and thirds. It is assuming a 64% turnout (not seen in any election going back at least as far as 1960) and lots of blacks coming out of nowhere to vote for Obama.

As I read their write-up, they further assume that turnout on election day will be 39D (with 90% party support), 31I, and only 26R. Yes, I typed that right. Oh, and Obama wins the Independents 48-43.

By the way, this is totally contrary to what Gallup has been saying in its weekly internals that I used to trust so much. Those party IDs were bad enough, but at least they measured registered voters and I was free to accept or reject them. Incidentally, the party ID for the last breakdown by party ID was

Dems 41.3 (4.3 oversample from 2004)
GOP 32.2 (4.8 undersample from 2004)
Ind 26.5 (0.5 oversample from 2004)

Anyhow, I really don’t have any response to all this other than to dismiss it and wonder of Gallup is on the take. Seriously.

The original 60% turnout model was predicated on a flood of new young voters and minorities (mostly blacks) who were going to vote overwhelmingly for Obama. Gallup has since admitted a few days ago that the yoots are NOT going to vote in significantly higher numbers than four years ago. Fine. I could have told them that months ago. But then why are they increasing their turnout model another 5%??

Answer: black voters, who supposedly are going to vote in so many numbers that they make up for all those yoots who would rather spend this Tuesday at the mall or whatever than voting for Obama. Gallup projects black turnout at 11 percent of the electorate. Well, guess what. Black turnout was at 11 percent in 2004! It’s true, check it out. Kerry won by big margins in big blue states with lots of black people. Now Obama might win by the same or even bigger margins in these same states. Or maybe not.

But it gets better. Gallup now projects that Obama will get 97%(!) black support from likely voters. That’s right: 97% of likely voters. [Insert “that’s racist!” here....but I digress.] And I’m now calling BS on this — I studied those Gallup internals every week and for weeks Obama polled consistently between 89% and 91% for blacks among REGISTERED voters (who are likely to be more Democratic than likely voters). Here, see for yourself:

http://www.gallup.com/poll/108040/Candidate-Support-Race.aspx

Which brings me to the hispanic vote. Gallup doesn’t make any claim that Hispanics will also turn out in record numbers from nowhere to vote for Obama. But it does claim in its final writeup that Obama is winning 73% of Hispanics. Again, I’m calling BS. Look at the link above. Obama scored consistently between 60-65% among Hispanics who were REGISTERED voters, not slightly more Republican likely voters.

The one good thing to take out of this load of horse manure is Gallup’s statement that McCain leads among white voters 51-44 — even with that totally nutty D39, 26R, 31I split. It is possible, even likely, that with a reality based split, McCain is much closer to Bush’s 58-41 performance of four years ago.

Also, I find it strangely encouraging that in a poll that oversamples Democrats relative to Republicans by a whopping 13 points, Obama still has only a 11 point lead, and even that only after Gallup allocated the remaining undecideds evenly.

[ubaldus]

>>What leads you to believe that the gap in party affiliation will go from +2 for the Democrats in 2006 to +5 or +8 in 2008? Do you really believe that the Democrats are that much more motivated to vote on Tuesday compared to Republicans?

1. In 2006 right track/wrong track differential was -30. Now it’s -65.
2. Definitely, it’s 2004 in reverse.

[rwilliam]

Here’s some historical data about the gap in party affiliation:

Year, Democrat, Independent, Republican

1972 46% 19% 35%
1976 41% 34% 25%
1980 45% 26% 30%
1982 46% 24% 30%
1984 38% 26% 36%
1986 40% 26% 34%
1988 38% 26% 35%
1990 37% 30% 34%
1992 38% 27% 35%
1994 36% 29% 35%
1996 39% 26% 35%
1998 37% 28% 35%
2000 39% 27% 35%
2002 38% 23% 39%
2004 37% 26% 37%
2006 38% 27% 35%

In order to find a gap of greater than +5 for the Democrats, you have to go back over 20 years to 1986. That’s right, even beyond the “It’s the Economy, Stupid” time period!

Sorry, but I’m not willing to throw out 20+ years of actual data about who votes in favor of unproven assumptions or feelings about who will vote on Tuesday.

-Bob

[snarkytart]

Thank you for that great analysis.
Come Wednesday, either Gallup or FReepers will look like fools.
I think it’s the former.
;)

[snarkytart]

PS

Now that Gallup has stolen the spotlight with this poll, I wonder if Zobgy will release a poll showing Obama ahead 15 to get some thunder back?

[cookcounty]

Gallup final estimate: O 55 - M 44

So if Gallup is as good as they were in the New Hampshire primary, Obama will lose by 5.

Or if they’re as good as they were in the California primary, he’ll lose by 10.

[ubaldus]

In LV, they have D 39, R 29, I 31 split. So, it’s D +10, and McCain running behind the Party ID gap.

Regarding black vote - they simply allocated undecideds to Obama. 97-1 is still a stunner, but 96-3 is entirely possible. +50 in Hispanic vote is very suspect, but I would buy a +35 estimate.

So, assume McCain is +14 in white vote (twice the Gallup margin, and btw R2000 poll by Kos shows him +14 there) and -60 in non-white vote (not -70). Then he loses by 5% or so.

Gallup chose to double down on its bets, though.

[goldstategop]

McCain will win the white vote better than Bush did four years ago, if the sampling had been fairly done. Obama is not going to win a majority of the white vote. He's not doing it even in Gallup's ludicrous final poll! So if he doesn't win a majority of white voters among men and women, how can he win the election by an 11+ point blowout? I'd really love someone to explain how that is possible, because no Democrat candidate in past elections going back to 1968 (except Carter in 1976) was able to replicate the feat. You can make a case for Obama being elected as a minority President, as Clinton was. But to give him 55% just takes it out of the realm of reality altogether. Gallup swallowed the Kool Aid in this poll.

"Show me just what Mohammed brought that was new, and there you will find things only evil and inhuman, such as his command to spread by the sword the faith he preached." - Manuel II Palelologus

[kesg]

Or as someone else suggested to me, maybe we will find out that someone had hacked the Gallup website. That final Gallup writeup had to be one of the craziest and intellectually dishonest things I have ever read.

[snarkytart]

For Gallups sake I hope their site was hacked.

[floridagopvoter]

Research 2000 is not really Dailykos’ poll. They are an independent polling firm, not affiliated with a party (like, say, Strategic Vision (R) or Democracy Corps (D) ) They have been around for some time now and were a pretty reliable state pollster in 2004 and 2006, as I recall.

I assume Dailykos is paying them for polls, but I would not automatically assume that to be a taint or a reason to doubt their polling methods. Of course it COULD be, but there is really no evidence to assume such, and the way they are weighting their most recent polls, and the findings they are arriving at (like many they are converging at around 5% to 6%) tells me that they are conducting valid research without undue partisanship.