How Far Ahead Is Obama Really?

by: Chris Bowers

Tue Oct 14, 2008 at 21:30


( - promoted by Daniel De Groot)

National polling is all over the map right now, with one poll showing Obama up 14%, another only 3%, and with other polls showing virtually every other number in between. So, how far ahead is Obama, really?

Twelve polling organizations have come out with likely voter polls in the last forty-eight hours. In an attempt to answer the titular question of this post, I took the most recent poll from all twelve, and averaged them together:

Likely Voter National Polls, 10/11-10/13
Org Obama McCain
ABC 53% 43%
ARG 50% 45%
Battle 53% 40%
CBS 53% 39%
DCorps 50% 42%
Dkos 52% 41%
Gallup* 52% 44%
Hotline 48% 42%
IBD 45% 42%
LA Times 50% 41%
Rasmussen 50% 45%
Zogby 48% 44%
Mean 50.3% 42.2%
Median 50% 42%

* = Only looked at the Gallup tracking poll, not the USA / Today Gallup. Also, averaged the two likely voter models. I figured, only one poll per organization.

The mean and the median matchup pretty nicely, both coming in at an 8% Obama lead. It also comes close to matching the Pollster.com trend (8.5%), and the Real Clear Politics trend (8.1%). It is a bit higher than the 538 trend, which shows Obama up only 6.4%.

In short, 8% seems like a very good guess for Obama's lead right now. 538's trend also seems reasonable--and certainly feels accurate to me, since I am a bit of a landslide doubter. However, there is also good reason to think it is a bit on the low side.

So, there you have it--Obama is probably up by 8%, but maybe only 6-7%. This has been another edition of reasonably simple answers to reasonably difficult questions.

Update: It appears that 538 also has Obama up 8% nationally, but the 6.4% is their current projection for the final outcome.

Update 2: Since I wrote this post, a new poll showing Obama up 51%-42% was released.
Chris Bowers :: How Far Ahead Is Obama Really?

Tags: , , (All Tags)
Print Friendly View Send As Email
How Far Ahead Is Obama Really? | 24 comments
Well it all depends (4.00 / 2)
On your partisan breakdowns and LV models, of course.  On the one hand you have Diageo/Hotline and, apparently, Zogby, with D+3 or thereabouts (Zogby may be even less - has anyone figured this out yet?), then you're looking at a 4 point lead or so.  Rasmussen at a D+6 shows a 5-6 point lead.  R2K and, presumably, some of these national polls that are coming out today are thinking D+9.  So it depends on who actually votes in what numbers on election day -- yes, the safe bet is to average them, but it could be one or the other end of the spectrum that is proven right in the end.

I think I made that far more lengthy and convoluted than it needed to be, but I'm too tired to revise.

-- Stu

538 (0.00 / 0)
For what it's worth, 538 also has the Obama lead at 8 right now, but they forecast it to narrow to 6 points by election day, and that's the figure they show in the left sidebar.
Beat me to it. n/t (0.00 / 0)

[ Parent ]
538 national average (0.00 / 0)
Well, 538 has two different national projections, one is a snapshot (in which Obama leads by 8.1) and the other is the projection (in which he leads by 6.4). It's the difference between the red line and the yellow line on their super tracker graph.
Ok Folks (4.00 / 2)
step right up, and pick a poll, any poll, and whatever one you pick it will show Obama ahead.

The time is always right to do what is right.
Martin Luther King
Well golly.... (4.00 / 3)
I'll pick today's CBS poll, if ya don't mind.  

"Don't hate the media, become the media" -Jello Biafra
[ Parent ]
In that new poll... (4.00 / 1)
Note that the w/3rd party numbers puts McCain at 39% again...

While Obama's lead is probably around 8% now, I wonder if we might actually see something we saw a bit of in the primaries now, where Obama's lead actually gets AMPLIFIED in the actual vote... McCain can't seem to climb past 44-45% in any poll now... will he be able to climb much further than that on election day if polls continue as they do?

I'm not predicting it, but suddenly a 10 point Obama win doesn't seem all that unrealistic.

9.7% (4.00 / 6)
That's the number I want to beat. That's what Reagan beat Carter by in 1980.

You really want to throw the fear of God into the GOP? You hit that number and you'll send a clear message to the Republicans that modern conservatism is dead.

Good point but (4.00 / 1)
what's another 0.3%? 'Double figures' has an awfully nice ring to it.

Gimme a 10!

[ Parent ]
That's Fine (0.00 / 0)
I don't care if Obama beats McCain by one tenth over 9.7% or ten percent over 9.7%, as long as he clears that bar and Joe Biden gets to go on the Sunday talk shows and say, "Barack Obama was elected by a larger share of the national popular vote than Ronald Reagan was in 1980." It would crush the GOP's spirit, to think that Barack Hussein Obama could actually out-do "white Jesus," (Boondocks reference) Ronald Reagan. It'd deflate morale for the GOP for the next decade, because they'd know that modern conservatism is dead.
[ Parent ]
not likely (0.00 / 0)
this election always has been and still is a close race. Obama by 5%. i hope i'm wrong and you're right, but i just don't see it happening. we still have 20 days left.
[ Parent ]
Strangely that Ipsos (0.00 / 0)
McClatchy poll seems to be still using RV and not LV.  I think they may be the only ones still doing RV only at this point.

-- Stu

Looking back at 2004... (4.00 / 2)
A commenter on pollster.com posted this big list of polls from the first half of October 2004. They range from Kerry +2 to Bush +5 for the outliers, but with most in the Bush +1 to Bush +3 range. That very closely mirrors the actual outcome.

There's also this analysis of the pollsters in September of 2004. They all showed pretty divergent results, but they converged on something roughly in the middle and all pretty much came close to the final result within the margin of error of the polls.

Take a look at the chart of the final 2004 polls. Such-and-such pollster may brag about being closest in 2004, but statistically, they were all tied, except perhaps Fox, which, ironically, predicted a narrow Kerry win.

If we take 2004 as a guide for how well the pollsters will do, taking their average is a pretty good bet. There were a lot fewer undecided voters in 2004 however. The obvious historical analog to this election is 1980, which could mean that Obama's lead is destined to keep growing.

Conduct your own interview of Sarah Palin!

If this distribution is bimodal... (0.00 / 0)
that is possessed of two distinct populations (and I think it is, one ranging between 2 and 6 points, and one between 9 and 14 points) then poll averaging doesn't make so much sense.

There are two peaks because of two sets of assumptions built into the polls.  One set is, implicitly or explicitly, assuming a normal turnout election.  The other is assuming a much higher (for Obama) turnout.

In order to assess the most probable outcome, we need to weight each distribution with the probability that the underlying assumption are correct (or use some equivalent factor).

Anyone have a good idea on how to estimate these probabilities?

This all seems too good to be true.... (0.00 / 0)
...I'm seeing a lot of McCain support in an area that shouldn't have any, and all the freepers are reporting their "Democratic" friends and people they know breaking ranks... (they talk about lawn signs and bumper stickers, too, but since Obama signs are hard to come by, I don't worry about those)... There was a right wing story about a leaked Obama internal showing them only up by 2 points in PA. Are the polls really oversampling democrats?  Are we getting set up for a huge fall?  The public PA numbers seem crazy, by every conceivable stretch of the imagination!

I guess I'm getting late night Democratic paranoia... It's just that the numbers seem way too good to be true, you know?  

REID: Voting against us was never part of our arrangement!
SPECTER: I am altering the deal! Pray I don't alter it any further!
REID: This deal keeps getting worse all the time!

Some Thoughts (4.00 / 3)
I've noticed a pattern in your posts over time here that seem to strike an unnecessary note of concern at times when it isn't warranted. Your above post, again, falls into this category.

Anecdotes of McCain yard signs or bumper stickers being everywhere, or some right-winger saying that nobody they know is voting for Obama, mean nothing to me. Show me some serious, hard data that backs up these assertions that this race is close on the electoral level. The empirical evidence, as documented by Chris here and Nate at 538 is completely overwhelming (I believe Chris made a similar point to another poster here who was voicing concern a day or two ago): Barack Obama is winning.

Actually, not only is Barack winning, but he is winning comfortably. This may be difficult for many of us to fully embrace, since were are used to losing Presidential elections in recent history, but the time has come for all Democrats, including those in the blogosphere, to show confidence in their candidate and his message. It is clearly resonating with Americans and pointing out this fact does not mean we are arrogantly dismissing the remaining work left to be done.

As for the malarkey about internal polls, all I can say is that evaluating the average of all public polling is by far the most tried and trusted method of electoral evaluation. Chris has mentioned this endlessly since his MyDD days and the results back him up 100%.

Furthermore, I am extremely skeptical that any hard internals of either campaign, especially Barack's, would leak at this point unless there was a motive (supporter complacency) to do so. Despite some mythology that internal pollsters are inherently more accurate than public pollsters, the data that is released typically from internal polling simply is not as transparent as public polling. I mean, how many times have people in the blogosphere written off polls (rightly so) released selectively by candidates or by campaign committees because they are internals?

Again, public pollsters provide transparency in their samples, as well as trendlines for us to evaluate, which makes their polling, when aggregated, every bit as good in my opinion as internal polling.

One final point I want to make is not to read too much into state visits by candidates and internal polling. On the day that the McCain campaign pulled out of Michigan, Barack Obama was campaigning IN Michigan.  

[ Parent ]
Thanks for that info.... (0.00 / 0)
...I needed that! :-)

REID: Voting against us was never part of our arrangement!
SPECTER: I am altering the deal! Pray I don't alter it any further!
REID: This deal keeps getting worse all the time!
[ Parent ]
To give a counter (0.00 / 0)
I've not seen any visible McCain support, but I have seen support for Obama.

For me seeing a democratic person even on the local ballot is rare, but I don't live in a swing state so that gauges enthusiasm on a non swing state.

I personally think that poll numbers are going to underestimate Obama's victory .

The liberal wiki
Send an email to terra@liberalwiki.com

[ Parent ]
Sticking with anecdotal information (0.00 / 0)
I spent the weekend in central MI (near Midland) at a family reunion at the rural home of my oldest sister.

We drove up from Flint.  In previous presidential elections, most specifically 2000 and 2004, Bush/Cheney signs were think as mosquitos on a warm summer night. This time, I think we saw a grand total of 3 McCain/Palin signs, and one hand-made sign for "Sarah & Todd Palin".  

I was encouraged because two of my Republican nieces (in their 30s) that had been very impressed with Palin when she first came on the scene, were now suffering serious "buyer's remorse". One says she will stay home on election day, the other is planning to vote for Obama because she got in an argument with some people at her church (she's born-again) when they told her she would be "abandoning her race and God" if she even thought about voting for Obama. She and her family no longer attend that church.  

"It sounds wrong...
     ...but its right."

[ Parent ]
Oh MI is great! (0.00 / 0)
The support is strong even in wingnut country... Ohio... not so much!

REID: Voting against us was never part of our arrangement!
SPECTER: I am altering the deal! Pray I don't alter it any further!
REID: This deal keeps getting worse all the time!
[ Parent ]
The Make Up Of The Electorate Is Still The Big Unknown (4.00 / 1)
How many of the new voters will actually vote?  How many are McCain supporters?  When these polls are being weighted how accurate are they being ore are they weighting them to support their side (e.g. Zogby)?

Will Obama's dominance in the polls suppress his voters?  McCain Voters?

I think the world of Nate Silver and find his call on Tampa Bay to be miraculous but he did underestimate their winning season by seven games (7.8%).  So I agree a final of 6.4% might be a tad low.

I think in the end we are looking at 7 to 9 point win for Obama.  That in itself is a pretty good thumping but I think the real victory will be in the states where he will best Bush's 286 win in 04.  Certainly topping 300 and just maybe hitting 350.

The Republicans really have no agenda other than being reelected and I believe this time around the people know it.

McCain is collapsing (0.00 / 0)
The bottom is falling out of this thing.

I would be surprised if Obama wins by less than 7 and not at all surprised if he wins by 10+.

I'm bracing for at least one more big scare (0.00 / 0)
Maybe tonight, maybe not, but it is coming. I can feel it.
[ Parent ]
I know I should be bracing for this (4.00 / 1)
because 20 days is a long time but I'm just not feeling it. It feels to me like we're going to get a calvacade of nontroversies that sputter out.

We shall see...

[ Parent ]
How Far Ahead Is Obama Really? | 24 comments
Donate to Open Left







QUICK HITS

Friends of the Earth thanks the OpenLeft community for the ideas you generate and your contributions to the progressive movement.
As an anti-spam measure, there is a 24-hour waiting period after registering before new users can comment.
blog advertising is good for you
blog advertising is good for you
LIBERAL INSTITUTIONS
STATE BLOGS
SEARCH

   

Advanced Search

Home | RSS/XML Feed | Register | Contact

  © OpenLeft 2007-2009, Powered by SoapBlox