Undecideds Collapsing

by: Chris Bowers

Mon Sep 01, 2008 at 13:00

Has there been a convention bounce for Obama? If so, how large? Right now, only three polls are available to provide a measurement. This morning's Rasmussen poll (August 29-31) shows Obama ahead 49%-46%, which compares to a 46%-46% tie in the polling immediately before the convention (August 23-25). CNN has a new poll out showing Obama ahead 49%--48%, compared to 47%--47% just before the convention. Gallup shows a larger improvement, as a two-point McCain lead in the immediate pre-convention polling has now turned into a six-point Obama advantage.

On average, this shows a four-point improvement for Obama, although most polls have not released their post-convention numbers yet, thus leaving the size of the bounce unknown. However, beyond the net increase, the most important shift that occurred over the past week is the drop in the number of undecideds. Pollster.com illustrates this shift nicely (more in the extended entry):

Chris Bowers :: Undecideds Collapsing

This is the highest point for both candidates in the national trend. The ten day trend, starting at 9/1 with 8/22 trend in parenthesis, is particularly illuminating:
Obama: 48.4% (44.3%, +4.1%)
McCain: 45.7% (42.8%, +2.9%)

Currently, only 6.9% of the electorate is listed as either undecided or voting third party, compared to 13.9% ten days ago. The last ten days, starting with the Biden announcement, has moved more than half of all undecided voters off the fence.

This is why Obama's bounce is larger than it appears at first glance. He actually has gained significant raw support, as he has equaled, surpassed, or come within 1% of his all-time high in every recent poll. At the same time, McCain has gained raw support, thus making Obama's overall net gains minimal. However, a 1.2% net gain is actually pretty significant when it is coupled with such a large drop in undecideds. A 2.7% lead with 6.9% undecided or voting third-party is actually much, much larger than a 1.5% lead with 13.9% undecided. In the latter, McCain must defeat Obama by only 10% among the undecideds in order to force a tie, while in the latter he needs a whopping 40% victory among the undecideds to vote a tie.

Relative to the number of undecideds needed to win the election, Obama's advantage has actually quadrupled. Further, he now sits only 1.6%, or less than 25% of the current undecideds, from reaching the 50% threshold. Still, further, Republicans will not make any convention gains tonight, as they have cancelled the prime-time speeches for tonight's convention. Although, given that Bush and Cheney were speaking, I was kind of looking forward to tonight as the easiest way to tie McCain to Bush that we could find.

Tags: , , , , (All Tags)
Print Friendly View Send As Email

Another day, numbers are steady (4.00 / 1)
Gallup today shows Obama keeping his +6 lead. Now it's 49-43.

Will the Republicans get a bounce out of their non-convention?  

Of course - that's the media narrative and the truth be damned. (0.00 / 0)
The MSM will claim that Obama retreated to Chicago and did nothing while the Republican Party (exploitatively) mobilized to put country ahead of (their usual corrupt, divisive, and destructive) politics and saved the day by personally laying their bodies across the (unfinished) levies to stop the biblical floodwaters.

Anyway, I just received a text message from the Obama campaign asking for donations - so I donated $10.01 to the Red Cross.

Help support "CRASHING THE STATES"--a Netroots Film!

[ Parent ]
My bet is (4.00 / 1)
If you look at the undecideds who were leaning towards and intending to vote for McCain they took a few looks at Palin and decided against.  

McCain always had the idea around him that he was socially moderate.  And in many ways Palin is a socially moderate pick.  But not in the ways those voters wanted.

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

Yes this is a good point. (4.00 / 2)
The mini bounce, much as its impossible to tell anything form day to day shifts, for Obama today, back to 49% in Gallup for example, may be exactly that, disappointment with Palin, and solidification of Obama's position.

Obama has been remarkably centered and strong in maintaining his message, his tenor and his overall campaign strategy, while McCain is playing catchup and respond and adapt, a very reactive campaign. Obama grows stronger and more clear and more Presidential with each passing day, McCain grows more wild, more unpredictable each day.


The government has a defect: it's potentially democratic. Corporations have no defect: they're pure tyrannies. -Chomsky

[ Parent ]
So true (4.00 / 1)
I can verify that by telling you several of my independant friends who were undecided have told me they are now voting for Obama. They state that Palin is just too extreme and inexperienced.
I really believed that this election was going to be won by who would get the independants and I now know that will be OBAMA!!!  

[ Parent ]
Obama moves back to 49% on Gallup as he maintains the 6% lead over McCain (4.00 / 1)
This retention of support, even some possible increase, as McCain announces a vp nominee, is very encouraging. It has to be remembered though that even the three day rolling poll is subject to number drift. Polling is a fuzzy science. The imgae of the daily trend movement on Gallup.com provides support that Obama's masterful excellent speech has solidified real support.

I anticipate exciting news as this national trend is matched in swing state polling. Which states move how much will be important, and I think exciting, polls. That's where we will get confirmation of this bounce, and of Obama's plans.


The government has a defect: it's potentially democratic. Corporations have no defect: they're pure tyrannies. -Chomsky

I've been reading the blogs for a few years now... (0.00 / 0)
... and have seen many references to the "50% threshold" that Chris mentions.  I understand that at the point where a candidate or a position gains 50% that somehow his/her/its chances of ultimately winning during the actual election goes up but why is that?  How much does one's chances go up?  Is there any statistical significance to gaining the 50% threshold?  Obviously it doesn't assure a 100% chance of the candidate and position winning, but what does it mean?

I'm sure that someone else has asked this question, so if anyone cares to at least post a link to the answers I would greatly appreciate it.


If you're at 50%, you don't need to convince anyone more (4.00 / 1)
If the poll is reliable and matches turnout, 50%+ support means you've achieved a majority. The other guy can't beat you if you keep all your supporters. He has to not just win undecideds but win your supporters. And that's much harder to do.

Forgotten Countries - a foreign policy-focused blog

[ Parent ]
The 50% rule (0.00 / 0)
primarily applies to incumbents.  When incumbents are below 50, they are considered in trouble because the undecided tend to break 2 or 3 to 1 against incumbents.

[ Parent ]
Except in 2004 (0.00 / 0)
Where most of the undecideds did not break to Kerry, resulting in a narrow loss to Bush.

[ Parent ]
2004 (4.00 / 1)
provides an interesting exception to the theory.

It was surprising.

[ Parent ]
interesting to track advertising behavior @ 49/50% (4.00 / 3)
I have found it very telling to track the behavior of the opposition campaign when Obama hits or comes close to 50%. Hillary and McCain both pulled out the negative ads and hit hard each time he approached or hit 50% in the daily tracking polls. Apparently it is a marker that instills fear in the opposition.

[ Parent ]
That CNN poll seems really screwy (4.00 / 1)
I don't think we can trust any poll two months out from the election that has only 3% undecided, do you?

My figures (0.00 / 0)
Here's what I just calculated, by averaging Rasmussen and Gallup from 8/24 through today.

In the immediate pre-Biden period (8/24-8/25), Obama was up by 1.5%.

In the post-Biden/pre-convention period (8/26-8/27), Obama was down by 0.5%. The announcement of Biden (everything else equal and all the other necessary caveats, of course) caused a 2% drop.

In the post-Michelle period (8/28-9/1), Obama is up by 4.8%. This gives Obama a net bounce of 3.3 for VP picks and Dem convention and a bounce of 5.3% for the convention and Palin.

Since Obama has been up by 4.5% for the last two days, it is probably more conservative to just call it an even 5%.

I wouldn't surprise to see this hold for a while what with the chaotic and reduced Repub convention; the thus far relatively restrained Gustav impact, which won't allow much of a sympathy vote; and the ongoing rapid implosion of the Palin pick.

I would avoid averages (0.00 / 0)
in bounce calculations where possible and stick to poll by poll comparisons.  In other Words, Gallup was tied and on Friday had Obama up 8.  Rasmussen actually had McCain ahead before the convention - I woul dput the Ras bounce at 4.  Between the two the average is six: which is interesting because the average post '96 bounce is 5.7.

[ Parent ]
I'm not ready to celebrate (4.00 / 2)
since I just read a report (I can't find it anymore, sorry!) that a large number of independents think Palin is ready to be president.  Bottom line - I want to see where we stand heading into the debates, and I genuinely hope that the Obama team has taken good stock of the Saddleback meeting to shore-up Obama's debate tactics against McCain.

The undecideds who have recently moved are moveable again, so we need to be really careful.


Visit the Obama Project


I'm pretty sure (4.00 / 2)
that was from the most recent CNN poll (but could've also been from rasmussen).

At any rate, I wonder if those independents were men who've were republicans who are now self-identifying as independents. So they were pretty conservative to begin with, and McCain choosing Palin brought them back. I doubt Obama would have won them anyway. But this is all speculation so who knows.

But what I do think has happened is Obama is starting to solidify Democratic women back to the levels he was at in late june early july when he got the post primary-bounce which then seemed to erode before the convention. That may be the best thing about the convention/Palin effect; moving women strongly into the Obama camp. Since Democrats and dem-leaning independents still far outnumber republicans and rep-leaning independents, then we should have an edge going into the election with this base solidification.

We'll know more next week after the dust settles from the dem-convention-palin-republican-convention-gustav-hanna-monday night football week.

[ Parent ]
There was also something released today (0.00 / 0)
showing that something like 2/3 of women respondents didn't think she was qualified. Or sumpin like dat.

Combine this with the rapid implosion of Palin and we've got reason to be cautiously optimistic.

[ Parent ]
Some may have been women (4.00 / 1)
Who initially went with McCain but go back to the dems when they realize how conservative McCain and Palin really are on social issues.

John McCain--He's not who you think he is.

[ Parent ]
Here are the number (4.00 / 2)
of undecided in polls taken after labor day.

I flat don't believe the number of undecided in recent polling: its one of the things that lead to my interest in state polling - which tends to show higher undecided.

One thing to note - if the undecided are really this low we should not see much of a bounce from the GOP convention.

Early Sept
Year Undecided
1984 6.3
1988 10.3
1992 14.3
1996 6.7
2000 4.4
2004 5.6

Interesting (4.00 / 1)
The fewest undecideds seem to be with reelection of a reasonably popular President--1984 and 1996, also 2004.  Highest are when there is sufficient discontent to force a change of party (including Carter's reelection loss)--1976, 1980, 1992.  It was high in 1988 as well, but that went to the R's.  The anomaly is 2000, when there was a change of party but fewer undecideds.  Maybe we are more polarized in the '00s.

John McCain--He's not who you think he is.

[ Parent ]
That is the conclusion (0.00 / 0)
I would make.  

As I have written before, the Convention Bounces have also decline since 1992 - they are about 60% what they once were.

In general the highest volatility happened in years where the right/track wrong track number was over 70% (1980 and 1992).  But that isn't the case this time.  

[ Parent ]
Who was the volatility between? (0.00 / 0)
1992 had Perot and my understanding is that Anderson posted some fairly high numbers in the summer in 1980. Would you happen to have the numbers for the highs and lows for the different candidates in these years? I'm just thinking that there could have been voters who liked two of the candidates but not the third.

Certainly it's been notable that McCain's national numbers have stayed within a relatively narrow band. It'd be interesting to see how much he's following previous candidates here.

Forgotten Countries - a foreign policy-focused blog

[ Parent ]
To round the season off we need to have the Republican Convention (4.00 / 4)
Right this moment, Biden did nothing, but didnt hurt as much as "anything but Clinton" was supposed to do, Michelle answered and provided real gains, Obama's speech pushed millions into the committed column. The Palin was announced to short excitement, but lasting questions. The sum total of which is a net week bounce to Obama of +/- 5 %

The only thing left is the Republican Convention, which may become a telethon for Gustav. How this plays out, with the MSM kissing McCain's feet, is what will determine that bounce.

The general scrum and scatter of Palin's candidacy dragging attention away from the convention is the unknown right now.

The huge plus is our candidate, and most recently the impressive speech given at the open celebration at mile high.

Right now thats what needs to be pressed forward. Not Palin, not Gustav. Obama's speech, Michelle Obama's and the Clintons' speeches and Barack's steadiness and readiness to govern.


The government has a defect: it's potentially democratic. Corporations have no defect: they're pure tyrannies. -Chomsky

Great, whoppee, yada, yada, yada... (4.00 / 1)
These national trending polls mean little, except how one party or the other can spin it to the big media-types.

What I want to see are the state-by-state polling effects of the convention/Palin pick.

This is an Electoral College election and the only useful polls are those that poll by state.  I mean, what if the largest block of "undecided" voters to move to Obama that is reflected in these tracking polls are actually residents of New York and were Hillary die-hards?  So what?  Obama was going to carry NY anyway.  In this case, all others states holding to pre-convention numbers, it was a wash, even though the tracking polls show significant improvement.

Just ask Al Gore how useful it is winning (leading) even in the last/only meaningful national poll- the actual election.

That will take a couple of weeks to show (0.00 / 0)
Lots to happen--the GOP convention, the Palin fallout, Gustav, then Hannah up the Atlantic coast next week, then Ike and Josephine just starting to form.  Lots ahead.

John McCain--He's not who you think he is.

[ Parent ]
The yadda yadda is the discussion on the issues you ask about in your post. (4.00 / 1)
I re read the posts, including my own, and see just the same questions asked already. The gallup daily tracking has, although atm i dont have the breakouts for, seem to presage the tacking in state by states that are, as you say, the vital actual elections that matter. I think point being, those 'undecideds' that Chris pointedly and thankfully identified as moving to one or another camp, have to live somewhere.

I anticipate exciting news as this national trend is matched in swing state polling. Which states move how much will be important, and I think exciting, polls. That's where we will get confirmation of this bounce, and of Obama's plans.
. . . is how how I yadda'ed it above.

The undecideds have to live somewhere. I don't think the great bulk of important newly committed will be from New York, I don't think its California, just because of the kind of voters and polling trends there. The swing states I think will confirm that a large number of people, who had early excitement, had their inklings confirmed by the Convention, by Obama's speech. I don't know this, and I have already apologized for not providing breakouts for you. But I think the coming single state polls will bear this out.

I'm waiting, yadda.


The government has a defect: it's potentially democratic. Corporations have no defect: they're pure tyrannies. -Chomsky

[ Parent ]
Yada, yada, yada wasn't meant as insulting (0.00 / 0)
to anyone posting here.

Rather it was a reference to ALL the discussion the tracking polls seem to generate.

I too think Obama generated some excitement and probably garnered many "undecideds" to his cause.

The two pertinent questions are 1) where are those new supporters located and 2) how solid (likely to end up voting) are they?  The daily tracking polls are useless for 1) and only tangentially informative about 2) (afterall, if Obama losses support on a daily poll are they the same people that recently converted or others, perhaps from a different state?).

So yes, if those new supporters are located in Ohio, Florida and Colorado, this is very significant.  On the other hand, if they are mostly in New York, California and Illinois (three of the most populous states and therefore statistically most likely to be the residence of the new supporters), so what?

So all the yada, yada, yada on these daily tracking polls is just that.  Empty speculation.

[ Parent ]
Thank you for your opinion (0.00 / 0)
I think however the understanding of political behavior is assisted by large number analysis, learning from people who understand it better than I is useful, to me as a citizen trying to understand the society I'm in and as a involved campaign to change the present office holders in Washington.

Again I dont think the larger number of, the greatest portion, or even the most significant portion of 'parked' Obama supporters, the ones waiting for permission to believe the wild idea that we can trust this inspirational and brilliant leader with our vote, live in New York or Illinois.  I think they are in purple and red states. Thats my analysis, based on thoughts found here.

Its an important insight, if true because it means Obama's methods and message is working; simple decency and direct, unstinting declaration of Republican/McCain failures is the right way to go. Early reminder of that could help everyone involved in similar campaigns to follow suit closely.

But I am willing to wait for the individual state polls to say I'm right. But in the meantime it appears being as direct and critical and focused on the American people is working. At least its working well enough to not risk temptation and act otherwise.

One persons 'Empty speculation' is another's way of finding compassion.


The government has a defect: it's potentially democratic. Corporations have no defect: they're pure tyrannies. -Chomsky

[ Parent ]
How does Palin impact Colorado? (0.00 / 0)
The national numbers are, ultimately, meaningless.  Despite Paul's hope for a new electoral map, it doesn't look like we're going to get one.  It's the same old blue/red divide this year and Obama has to win all the Gore and Kerry states plus at least one.

I'm thinking Palin is targeted specifically at Dobson's evangelical infrastructure in Colorado, the state which perhaps the GOP had seen as likeliest to flip.  (Also at Ohio, though the christians haven't been so successful there just lately.)  Colorado, Ohio, Virginia, Florida, Nevada?  Which one can we win?

The collapsing of all the undecideds toward a close race again seems to indicate that cheap last-minute emotionalism, electoral dirty tricks, and perhaps some October military surprises will be determinative of the presidency.  (To my mind, it's also evidence that our democracy is completely dysfunctional and trivia-based.  We've got a serious long-term problem to solve, no matter who wins.)

I think they may think (4.00 / 1)
she helps in some of the small western states that have looked close.

This election is close: holding all of the small western states matters.

[ Parent ]
that assumes there are more votes to be mined (0.00 / 0)
there. mccain's problem never was his base. if it were he wouldn't be stuck with the numbers he has. its everyone else. i don't have any proof of this, but i will bet if you look behind the numbers in the states-- its not that he's losing the far right vote that's killing. its that he's losing the independents and moderates. how does she help this situation?

[ Parent ]
Paul, apologies for being egregiously off topic (4.00 / 1)
But are you the author of this:

The Empire Strikes Back:
Police Repression of Protest From Seattle To L.A.
By Paul Rosenberg
Written For The LA Independent Media Center
August 13, 2000

note: it's a pdf

I haven't read enough yet to decide is it "sounds" like your "voice." But, decided the most reliable way to know, was to ask.  

Yes, That's Me (0.00 / 0)
Hmmm.  I've still got a copy of that.

But I don't have a copy of the series I did shortly after that, about all the racist talk show hosts who'd won national broadcast awards.  And the IMC site I did it for wipes itself out from the archives.  So if Philly didn't... you may have just given me a valuable lead!

"You know what they say -- those of us who fail history... doomed to repeat it in summer school." -- Buffy The Vampire Slayer, Season 6, Episode 3

[ Parent ]
sorry, chris! (0.00 / 0)
got tangled up when I logged in.  not only off topic, but inappropriately so.  my bad!


Open Left Campaigns



Advanced Search

Powered by: SoapBlox