Why is Gallup giving Romney such a large lead?
Poll attracting attention for breaking from the pack
- Duration 8:43
- Date Oct 21, 2012
Poll attracting attention for breaking from the pack
Also in this playlist...
This transcript is automatically generated
The Gallup organization has been polling in presidential races since 1930 sex which is why its latest survey -- attracting a lot of attention.
Gallup seven day tracking poll now shows Romney leading Obama 51% to 45 among likely voters that's a six point margin.
In comparison the real clear politics average of -- poll shows identity with the Obama leading by 110 of a point.
Joining us now from New York is Frank Newport Gallup editor in chief frank how do you explain the fact.
That -- poll is so out of line with seven other national polls would show Obama with a small lead or Romney with a lead of just one point.
Well there are many different explanations -- that we spend most of our time Chris looking at our own polls in our own methodology how we're doing and it's a seven day rolling average the one you quoted there is through last Friday a large sample size.
3500 registered voters 4700.
Likely voters in that sample we don't spend a lot of time looking at the methodology or how other polls might be done but what we can say is that we've been here a week ago.
We would have said the race was dead even ourselves and our Gallup poll would during this last week night by night with are fairly large samples we aggregate together we've seen -- move out to that the 5145.
Seven day average that we talk about and that's what we're showing among likely voters nationally registered voters it's -- much closer to being either.
You also had a poll this week of twelve battleground states and let's take a look at that it shows Romney leading by four points in this is in the twelve battleground states.
Fifty to 46 and that the gender gap among women has gone.
Obama leads by only one point 49%.
To 48 again.
Most polls show Obama leading by more in the battleground states and with a much bigger -- Among women.
But that's right now that was a couple of weeks ago actually we -- -- work was done -- think that's to some degree ancient history as I mentioned of these things changed quite a bit when we go week to week.
Nationally now in the aforementioned large sample -- there were talking about Obama does have a lead among women actually a seven point lead among women and which is certainly down from where it was -- -- But nevertheless he's maintaining his strengths among women.
It looks like there was isn't his bigger problem Obama -- -- is that he's lost more among men than where it was back and only.
Now they the Obama camp not surprisingly is pushing back hard they say that your polls.
Our quote way out of line and have deep flaws -- respond.
To the Obama campaign.
Well that's not -- going back to doctor George Gallup who founded our company few minutes -- -- -- 1936.
He found that he did commentary shall we say from either side.
Now on polls and we certainly found -- in this six election cycles and I've been involved here at -- going back to 1992 that people come at you from either -- if they don't like the results.
We do we do not have deep flaws in how we do things were very highly transparent have a team of methodology distance scientists to -- best in the business.
Constantly looking at what we do and how we do sampling we -- sampling we tweak.
How we interview we just added to actually more cell phones to try to take the into account of growing number of people who don't even talk on -- -- of these phone samples.
So I think our methodology it is extremely solid.
We're very open about how we do it so I would say you over to integrate and is not -- of course the people would fight back after the don't like -- fine.
Well let's talk a little baton and maybe get into the weeds -- people seem very existed in polling methodology this time.
The big complaint that is made about the Gallup poll is how you judge who's a likely voter and I say you what.
Too much emphasis on how enthusiastic.
About -- as and that therefore you may tend to overstate.
The effect of a big event like for instance the first debate.
Which may have boosted one side in this -- Romney's supporters.
And depressed and other side in this case Obama supporters but about this argument that there's too much.
And therefore we see these big swing was based on enthusiasm or intensity based on events in the -- and at in the campaign.
Well you know how you isolate likely voters is a very fascinating met the logical question that we Dugard -- and most of the polls do it actually enthusiasm is and one of the questions that we ask when we measure likely voters -- -- seven questions.
That we've tweaked over the years but they include questions about knowing where people vote or how you vote if you go by mail.
How much attention you're paying to the campaign -- -- thought you've given to -- how certain you are devoting your own self definition in your history of voting.
Which we -- take into account for young people who couldn't voted previously we put all that together.
And isolate likely voters it can certainly be susceptible to event to -- environment -- that's a whole idea.
If events in the environment -- one group or the other urge to become more likely to vote based on these measures that's what it reflects.
-- you know when we have Chris is a democratic incumbent.
As we do now running for reelection though likely voters make a big difference back in 1996.
I had all voters voted based on our estimate -- woodcock Clinton would have beat dole by like fifteen points and of course the only one in the high single digits that's because the turnout would disproportionately for Republicans.
In back in 1980 we had another Democrat running.
And of course when the dust settle their Reagan -- significantly more than actually we're anybody else -- said to the Republican for more energized so I think likely voters trying to take into account.
All of these factors that I just mentioned about who's going to show up and who ridden.
Is extremely important particularly election like this which actually may have lower turnout than we saw no way.
Well let's talk about the other big controversy -- this isn't just for you which are all pollster is and that's the question of how do you wait.
-- party what percentage of Republicans what percentage of Democrats on people's ideas in 2008 model.
As to how many what the percentage of Republicans and Democrats -- -- -- -- in 2010 model which obviously is different how do you -- -- part.
We do not -- -- party at all we never have and we don't now.
We think party identification is an attitude and of variable that fluctuates just like you're gonna both -- to we let that be floating.
Whatever people tell us -- the end of the survey when we say as we say it gallop as of today he do you consider yourself a Republican and independent or Democrat.
That's what we do is measure at the end for informational purposes but we made no effort to try to wait by.
Federal and there are no national number dorm party identification the census doesn't measure party identification the exit polls and all later flog themselves because -- -- poll.
It certainly nobody thinks that nothing will changed over four years you don't have to register in some states there's no national registry of what party -- quote should be.
We do know what age should be in gender of the senses measures that we don't know.
And frank let me considerable weight that party -- -- I -- that this -- the -- the critics ask.
If we see Apollo we say well it's 39 present Democrats in the -- 30% Republicans that doesn't necessarily mean that it's it's my -- spewed.
Now I mean you can have that's like saying you see a poll that shows Obama up by nine points over Romney is that bias or skewed well it may be that a particular poll has a spurt.
As we call it in the business that is you know margin of error an unusual one way or the other but usually if the ballot is going to one directed the party had we will follow right along behind it.
Again party idea just as much.
In many ways an attitude says who're you gonna vote for or so we certainly don't think it actually most of the major national polls do not wait by party I'd be because that's -- at -- a -- have a lot.
Science -- up front with a little bit over a minute left as we look at this raise sixteen days out.
Where do you say it.
Well we don't make predictions here at Gallup and again I told do a week ago we had an even and now we have -- -- -- but that movement suggested it could move back the other way over the next week or two as lousy as of today certainly Romney in the national level has a marriage that's what our data are showing.
All the swing state polls and all that individual states are very hard to make sense of that they look close themselves that I think.
Right now I wouldn't predict I would say we got two weeks ago and a big debate tomorrow night and.
And it just looking historically over the long history of Gallup polling and 86 point lead which is what you have now sixteen days before an election how solid is that.
Well not solid at all in the sense that.
Things can change it change from a week ago and we might talk again next -- -- nice to talk with your -- and we could see where we stand there and it could be significantly different things move -- election that's why they're still spending.
Hundreds of millions of dollars on both sides to try to move voters you know the campaign hadn't stopped it's in full steam ahead at this point.
Frankly want to thank you thanks for talking with us today and we'll be following -- poll and the others all the way to election night.