Gallup 10/26 Romney 51 Obama 46 LVs

[dennis4868]

[fugop]

1. They're all in on that crap likely voter model

Because then they'll have the excuse that their likely is off due to its ridiculous assumption that the electorate will be up 7 percent in GOP from 2008.

[Johnny2X2X]

2. Utterly insane

Wow, not good. I'd rather not be left to tryinig to say polls are garbage, those people ususally are the losers.

[Mutiny In Heaven]

4. Not all polls.

But Gallup is one of the very worst. A horrible track record spanning several decades now.

[fugop]

12. People need to give up on Gallup's likely model.

As Obama's people said, it's a crappy model. And if it's not, nothing will win the election for us.

But it is. A crappy model, that is. Their registered voters are closer to reality.

[regnaD kciN]

3. Ick...

That meant Rmoney gained two points -- must have had a good polling day (as opposed to the ABC/WaPo results).

[mgcgulfcoast]

5. getting depressing

[Mutiny In Heaven]

9. If you consider Gallup an accurate polling firm

which history tells you would be incorrect.

[russspeakeasy]

6. I heard it was rmoney 97 Obama 1.

The asshat pollsters are looking for attention.

[aaaaaa5a]

7. According to Gallup Obama had a really bad polling day

his approval is -3 and is now 48%

[tbennett76]

13. They are 10 pts

off on the makeup of the electorate. They are projecting a 2004 model and they are not even close.

[woolldog]

16. do you know which day?

doesn't look like they're going to get any better

do you think Libya might be weighing him down? The media's been harping on it constantly esp Fox and CNN

[aaaaaa5a]

19. No. But I can guess

Its one of two things.

Either Obama had a bad day of polling yesterday.

Or he had a good day of polling that fell off and Gallup's numbers returned to its normal means.

When looking at LITERALLY EVER OTHER POLLSTER OUT THERE it looks like Obama had a good day in the model that fell off. If he truly had had a bad day, it would have been reflected in other tracking polls.


Given Gallups flawed demographic composition, this looks like this is there base assessment for the race. And given the polls makeup that would be a correct analysis.

The problem is the composition is wrong.

Gallup Numbers for their National Poll

W-78%
B-11%
H-7%


2008 Final numbers

W-74%
B-13%
H-9%


Likely 2012 Final numbers

W-72%
B-13%
H-10%


If whites are 78% of the electorate on election day, Obama will lose. And he will lose by about the margin Gallup predicts. But NOBODY believes the white share of the electorate will be 78%. 72% is a more realistic number. If you plug the 2008 or 2012 demographic profiles into Gallup's model, you get 1 or 2 point race, which is where the Obama/Romney contest really is on a National level.


On a state level, the President continues to hold an advantage. And in fact his map is beginning to expand. Colorado was thought to be leaning Romney. Today it is a tossup to lean Obama. NH is tossup to lean Obama. Virginia is exactly tied. And Obama holds leads in OH, WI, and IA. Nevada is no longer a battleground state. It belongs to Obama.

This race is actually shaping up very much like Nixon/Humphrey in 1968. (There's an analogy you haven't heard!) The popular vote between Nixon and Humphrey was very close. But in the electoral college Nixon dominated. This year we could have a 1 or 2 point race nationally, but Obama basically declared the 270 winner around 10pm EST when IA is done counting ballots.

Remember regardless of any National poll, If Obama holds Ohio and Wisconsin he wins the Presidency. Nothing else matters.

[woolldog]

20. thanks. but why do you feel whites will make up 72% of the electorate as opposed to 78% ?

How do you have such a strong opinion on that?

[DemocratSinceBirth]

22. Because The Non White Percentage Of The Population Is Growing Much More Rapidly.

The Latino population is growing faster than the African American population and the African American population is growing faster than the white population. That's simple arithmetic.

[aaaaaa5a]

23. Because the White share of the electorate has been going down for the last 30-40 years.

Here is the percentage of the white vote for every presidential election since 1976.



1976-Whites 90%

1980-Whites 89%

1984-Whites 87%

1988-Whites 85%

1992-Whites 87%

1996-Whites 83%

2000-Whites 81%

2004-Whites 77%

2008-Whites 74%


The white vote on average drops about 2 points every 4 years. In fact from 2000-2004 it dropped 4% points. And from 2004 to 2008 it dropped 3% points. This is critical. So critical that as dominate as Obama was in 2008, if the electorate had looked like it did in 1976, McCain and Palin would have won!

No Democratic President has won the majority of the white vote since LBJ in 1964. In fact when Bill Clinton won in 1992 he had just 38% of the white vote.

Based on these projections and the current growth of hispanics, the white vote should be around 72% of the electorate in 2012. It could be as low as 71%.

Gallup's poll numbers assume a white vote of 78%! Which as you can see given the current demographic projectory of the country, is way off base. This is why Gallup's numbers are such outliers. It is really too bad the MSM can't do their homework so they can report figures and context accurately.

I honestly don't know what do make of this country when an average "Joe" like me sitting at home on a Friday afternoon doing paperwork and messing around on his laptop, literally knows more about these numbers than CNN.

[woolldog]

24. per Gallup

whites were 78% of the electorate in 2008 and will be 78% of the electorate in 2012. It does seem strange that the number would be the same as in 2008. But I have no way of knowing what the correct number should be. I suspect it's tied in with their LV model. Do we have any idea what other pollsters are finding?

[aaaaaa5a]

25. Actually what Gallup was saying was that

in their final poll of 2008, their polling demographics were 78% white. However the 2008 voting exit polls were 74% white, which was right in line with the declining white vote compared to exit polls of 2004 and 2000.


Gallup is trying to justify the faulty demographics in their polling by saying it is the same model they used in 2008.



It is factual general knowledge that whites were 74% of the electorate in 2008. Even the folks at Gallup know that.



What you posted were Gallup's 2008 likely voter models. Not the exit poll results.


I would add that it also shows how faulty their likely voter models really are. They said in 2008, 78% of likely voters would be white. When in fact only 74% were,

[mvd]

8. Don't put any stock into it, especially with non-Gallup and Ras polls..

Tightening today.

[Maximumnegro]

10. Romney will not get below 2% on Gallup. Period.

See WI Dem's other thread. Stop watching Gallup. You will only depress yourself. Those numbers are not going to come down. It's what they do.

[tbennett76]

11. You can add 8

to this poll for POTUS the current state of the race is Obama 50-47. Both sides agree on this BTW.

[timlot]

14. I voted today...

And I can guarantee you I'm not counted as a likely voter. We'll see how things turn out I can't wait for this to be all over.

[leftynyc]

15. For what it's worth

This came from Daily Kos with a link to gallup

Although, the demographics look almost identical in 12 as in 08, it seems Gallup in terms of party id are using '10 numbers.

In 08:

Among RVs: 39D, 29R, 31I
With leaners: 52D, 42R

In 12:

Among RVs: 35D, 36R, 29I
With leaners: 46D, 49R

I don't like skewing/unskewing polls. Perhaps they are really finding this party id data with their polling, but at least we now know how they are getting their numbers. Personally, I find this highly unlikely but what do I know.

Link: http://www.gallup.com/...

[Mutiny In Heaven]

17. Why is no-one else finding that?

Are they avoiding cell-phones and intentionally refusing to weight the poll accordingly?

[ProSense]

18. Gallup is becoming a joke.

Gallup: 2012 electorate looks a lot like 2008
http://www.democraticunderground.com/1251179150

The polling model is clearly off. How the hell can there be such huge swings in approval and likely voter numbers from one day to the next?

[iloveObama12]

21. U Might As Well Say Romney/Gallup Poll 2012! What A Complete Joke This Organization Is!