while the other polls are seven days. In the last three days the job numbers have improved. It could very well mean that on the horse race numbers that we are on our way to see better times, too, once some bad days over those seven days drop off.

But Gallup has never shown such huge numbers for Romney before.

So the question still remains, WHY?

but then shouldn't the LV number also be going down, since it added yesterday (which accounted for a 1/3rd of the job approval number)? It doesn't really make sense for them to be going in opposite directions. The LV number as a lagging indicator makes sense, but for these to be in that much contrast does not. There is a 13 point difference.

Can't even get mad at this point. Too hilarious.

That is right wing indeed! They have now reached Gravis credibility levels. Hope they are challenged as to why their poll is so different than the rest.

...it sounds like such a nice scenario for voting machine tampering to put Mittens in the White House. If there are any exit polling done, these kinds of fake numbers will be the oppositions' talking points.

You can't bloody well steal the election without first getting the polling in-line with your "election plan".

look at this IBD pollster. they arent too famous but they were 100% accurate in 2008. They predicted Obama ahead of mcCain by 7.2% just hours before the elections, and Obama won by 7.2%!!!

Gallup's CEO is mad that Obama called them out on their swing state outlier, so he's trying to retaliate. Either that or their LV screen is a joke.

.cms

That likely voter model is a big laugh and shows their rightward drift.

Note that even without the LV screen, there are still millions of people saying they approve... but will vote for the other guy.

Approval rating just isn't the firm predictor of vote share that some people think it is.

Rand, Ipsos, IBD.

Gallup has Romney +22 in the South and Obama +4 in the East (CT +12, DC, DE, MA +19, MD +20, ME +14, NY +26, PA, RI +22, VT +28)


There is no way that Romney has a 1-point RV lead and a 7-point LV lead with the President's approval going up 2 points.


Absurd!

... yet still plan to vote for him?

For starters, yeah, I'd wager most of those people who don't approve of Obama will probably vote Romney. So, in your scenario, it wouldn't be shocking ... considering that's generally how elections work. Yeah, there will be a few, on both sides, who approve/disapprove of the President and vote for him/or his challenger despite logic ... but that is generally a small number and not enough to dramatically change the election in any way. Here, we're working with a five-point swing in their LV model. FIVE POINTS. I call bullshit.

It clearly IS possible to disapprove of a candidate, yet vote for him anyway (many here on DU fall into that category)... it's just as possible to approve of a candidate, but prefer his opponent. I don't think anyone is polling Romney's "approval rating", because it's really a job approval question and he doesn't have the job.

Here, we're working with a five-point swing in their LV model.

Which says more about voter enthusiasm than approval. Debate the above paragraph all you like, but we KNOW that's it's possible to approve/disapprove and simply not care enough to show up and actually vote. See 2010.

But the gap we're discussing is really between approval and RV support... and there simply isn't a tight correlation between the two. Moreover, there isn't a historical corrective metric that allows a pollster to refine their methodology. You can align reported voter intentions with actual poll results... but you can't validate whether or not people who SAY they approve/disapprove actually do so (or mean the same thing when they say it). Just look at the variation between polls that ask it as a binary question and those who have four categories.

Then show me all these presidents who've lost reelection with approval at, or above, 50%?

There simply isn't enough data to state that approval above a certain level gurantees a win (and there's certainly nothing saying that approval below a certain level means a loss).

Gerald Ford was right around that point (45-50% in Mar-Jun and 53% in December), but Carter won by two points.

Who are all these presidents with majority approval who've lost, especially by seven-points nationally? Even your own example proves my point. Ford had an approval rating of roughly 45% on election day (I have yet to find a poll that puts him at 50 or better ... the only polls that do that are post-election polls which, traditionally, are more generous to losing candidates) and still only managed to lose the election by two-points and a very slim margin in the electoral college.

The presidents who've lost reelection have all had approval no better than 45%:

1. H.W. Bush: 34%
2. Jimmy Carter: 31%
3. Gerald Ford: 45%

Candidates who most certainly would've lost reelection had they run:

1. Johnson: 43%
2. Truman: 32%

Presidents who've won reelection:

1. Bush: 50.2% (an average of approval polls between October 28th - Nov. 2nd)
2. Clinton: 54% (an average of approval polls between October 29th - Nov. 2nd)
3. Reagan: 58%
4. Nixon: 57%

Obviously the better approval means the better electoral landslide. Clinton, Reagan and Nixon held better approval ratings, but won by landslide margins. Obama's approval is not much different than Bush's ... who won reelection.

You say there isn't data to prove my point ... but there is no data to suggest it isn't a predictor. The fact that no presidential candidate in modern history has lost with an approval within only a handful of points of 50 shows me that there is a direct correlation. The only candidate to come close was Ford, who trailed by huge margins throughout the '76 election ... had to fight back his pardon of Nixon and was running to keep his party in control for a total of TWELVE YEARS. In any other election, Ford's approval might've been able to get him that election.

But alas, as I said, even then, he still only lost by a narrow margin.

Gallup is telling us that Obama has an approval of 50% ... but that he's losing by seven nationally. Even H.W. Bush, with an approval of 34%, only lost the popular vote by 5.5 points.

And we're to believe this number is right?

Eisenhower was between 68% and 75% approval on election day... but only received 58% of the vote. Lyndon Johnson had 74% approval in the last Gallup poll before the election... yet pulled 61% of the vote.

Bill Clinton had a 58% approval (Gallup) on election day yet received only 49% of the vote. That's enough to win in a 3-way race, but it clearly proves that you can be well above 50% and still not have all of those "approvers" decide to vote for you.

Bush I was not at the 34% you cite. His final gallup approval was 43%... yet he only drew 38% of the vote.

That's four clear examples of incumbents under-performing their approval number by more than the current gap in the poll.

Just give it a few days,the 7 day rolling average can swing wildly in a couple days. If Obama is polling at a 50% approval he's ahead and the 7 day average will eventually show that.

given that even rasmussen only has Romney up by two points. And you have a couple of other tracking polls that show the President up by a couple of points. However, I have a feeling that romney probably does have a slight lead currently at the national level. If you look at the latest state polling things have become awfully close in Ohio, as well as some of the other swing states. Good thing is we're still almost 3 weeks out from the election, so it wouldn't take much for President Obama to flip those numbers. Whatever the case, I certainly don't believe Romney has anything close to a seven point lead nationally right now.

So, 50% of the electorate approve of the President's job, but only 45% are willing to re-elect him?

5% of the electorate wants to replace a President despite approving of the job hes' done?

Am I reading this right?

They have to be in the tank for Robme.

Yesterday Daily Kos showed the internals on the Gallup polls where the South got weighted preferably in their internals:
http://www.dailykos.com/story/2012/10/17/1146199/-Gallup-Poll-Geographic-Weighting

_______________Obama Romney Margin

.
East___________52 _____48______ O+4

.
Midwest________52 _____48______O+4

.
South __________39_____61______R+22

.
West___________53_____47______O+6

could the gap between LV and Approval have anything to do with our President being Black?

i'm not trying to stir up a race war, but for God's sake...prior the Bin Laden get everyone was like "if he gets Bin Laden...then everyone will love him" yet in the days AFTER he got Bin Laden his approval level was only at like 67%.

so maybe the same can be said for Likely Voters? "ya well he's done a decent job as President, but lets give the white guy a shot now".

maybe i'm just over cynical.

because I moved between elections and I would have to answer the question, "did you vote in your precinct in the last election" with a No. Therefore, I'm not a likely voter. Even though I vote every primary, every election.

So there's that.

If you voted in the primary, you would get a perfect 7/7 on their screen (because the question isn't "did you vote there in 2010?" but rather "have you voted there before?&quot . If you moved in since the primary, you would still get a 6/7 as long as you have an idea where to go to vote. Both pass the LV screen.