Back To A National Toss-up

by: Chris Bowers

Mon Aug 04, 2008 at 12:35


Nationally, the Pollster.com trend has moved to a toss-up. Obama is still projected with a small lead (46.6%-44.4%), but with less than a 68% chance of actually being ahead. There are three very interesting items to note about this trend:
  • Obama's numbers haven't dropped, and are actually on a slight uptick. Instead, McCain's have risen quite dramatically, gaining 3.2% in just one week. His new identity based lines of attack appear to be consolidating some undecided in his favor, rather than peeling off Obama supporters. It isn't about gas prices or drilling--coverage of that issue has been dwarfed by the identity-based attacks.

  • This is still a trend that has only appeared in two polls: Gallup and Rasmussen. While you can't ignore the Gallup and Rasmussen tracking polls, they provide an incomplete picture on the state of the campaign. Other polls have to confirm in order to prove the trend.

  • What changes there have been have only appeared in the national tracking polls, not the state polls, where my forecast has remained static since July 24th (there haven't been any state polls to change the forecast since the last update). At the same time, Obama's brief, post-trip bump never appeared in the state polls. He caught up in Florida, where he is spending more on ads than any other state, while McCain isn't even on the air. By contrast, McCain closed the gap in Colorado and Ohio during July. Other than that, there really hasn't been any change in state polling since June 26th.

Which is to say that while things are not great right now, the campaign might not be as mobile as it appears when looking at only Gallup and Rasmussen. Not only are Obama's numbers not actually down, but we have to wait for more polls to confirm at the national and statewide level. Still, there is good reason to except that when those come out, they won't be great for Obama.

The new presidential forecast will be up in a few minutes.

Update: Obama leads by 46%-43% in today's Gallup poll.  

Chris Bowers :: Back To A National Toss-up

Tags: , , (All Tags)
Print Friendly View Send As Email
Back To A National Toss-up | 22 comments
What do we know about their samples? (0.00 / 0)
 A recent NC poll was based on the assumption that African-American turnout would be DOWN 4 percentage points from 2004. Are Gallup and Rasmussen using similar turnout models? If so, it would be hard to take their results seriously....

Also, are these polls taking into account the huge shift in partisan identification since 2004 (Dems up 700,000, Repugs down 1 million)?

my memory of 2004 (0.00 / 0)
is that there was a time lag between when Kerry started dropping in national tracking polls and when it started showing up in state polls. Then, when he had a boost after the first debate, there was another lag before he seemed to be gaining in state polls.

I fear Obama is in for a rough round of state polling the next few weeks, but the electoral map still seems to favor him. I still don't see which Kerry states are at serious risk of going for McCain.

Join the Iowa progressive community at Bleeding Heartland.

Agree (0.00 / 0)
Tracking polls are often leading indicators of state polls, although forming an aggregate picture of state polls to detect overall shifts among them is hard.

However, I think there is a possibility that Obama's state polling - at least in battlegrounds - will outperform his national polling.  My guess is that with a candidate like Obama, there will be a lot more deeply red states than deeply blue states.  We are going to see some places in the south and appalachia go 65-35 for McCain, but even liberal strongholds like Massachusetts are likely to only go around 57-43 for Obama.  With Obama, there seems to be higher than usual chance of winning the popular vote narrowly but the electoral college quite widely.

John McCain: Health insurance for low income children represents an "unfunded liability."

[ Parent ]
Probably (0.00 / 0)
It is likely that future national and state polls will follow the Gallup and Rasmussen trend, but it isn't a guarantee. It is best to wait before jumping to conclusions.  
[ Parent ]
The lag theory (0.00 / 0)
Isn't it also possible that people are looking for some way to connect the daily tracking polls and static state polls, and that they look for random float in the daily tracking polls as evidence of a lag in order to reconcile the two sets of polls?  In other words, if the margin of error for a daily tracking poll is 3 pts., and Obama is tied in the daily tracking poll but ahead by 3 in a state poll, and then a few days later the daily tracking poll shows him up by 3 again, would that actually be evidence of a "lag"?

Maybe I can word that better, but I wonder if there isn't a perceptual error at work here.  

[ Parent ]
there isn't necessarily a connection (0.00 / 0)
I think I see what you mean. And yes, random fluctuation in the margin of error in the national tracking polls can account for stability in the  state polls.

It can also be that we don't know where the national numbers are changing, and they might not change in concert with the state polls.

In other words, changes of less than four points (they have a margin of plus or minus two) in the national  tracking polls should not be taking as a sign of pretty much anything. And, it is always best to look at the entire polling picture, not just the tracking polls.  

[ Parent ]
Olympics (0.00 / 0)
I wonder if the Olympics will put Americans in a more international flavored, Obama friendly mood?  Probably just wishful thinking on my part, eh?
I kinda doubt it? (0.00 / 0)
Isn't much of the olympic coverage going to be tinged by negativity stemming from the spotlight the olympics are focusing on PRC totalitarianism?
[ Parent ]
Because of the lag (0.00 / 0)
We might actually see some good state-by-state numbers before we see some bad. Maybe by then though the conventions might mitigate that somewhat.
Also as a pointed out in the Quick Hit (0.00 / 0)
This might have been affected by the fact that Ras found less Dems in July so he has adjusted his August sample accordingly. If we were still in July the same numbers may have still resulted in an Obama lead. Of course that doesn't change the fact there were less Dems in July but I think it is worth noting just the same.
[ Parent ]
Gallup +3 Obama (0.00 / 0)
I give up!
Yea! but it's nothing to shout home about either. (0.00 / 0)
Obama vs. McCain

and we're up 3 points...

[ Parent ]
Let me guess your response if he was up 10+... (0.00 / 0)
..."but Dukakis was up 17 points and lost!"
[ Parent ]
After 8 Years of Bush (0.00 / 0)
... and the invention of the Internet -- yes I was hoping for "little" more than 3...
[ Parent ]
Too close for comfort I agree but... (0.00 / 0)
...the reasons why have been pointed out ad nauseum.
[ Parent ]
Well then! (0.00 / 0)

[ Parent ]
Gallup... (0.00 / 0)
Does a rolling average each day of the past three (or five, I forget exactly) days of results.  Because of this, you'd be hard-pressed to have a net gain greater than say 4% for either candidate on any given day, just slow creeps upward and downward, generally within the MOE.  
[ Parent ]
Expectations re: crushing McCain (4.00 / 1)
People need to NOT believe that Obama should somehow be crushing McCain, and then base their criticism on that belief.

That is emotionally based, and I've seen it pop up in a few threads recently.    

What? (0.00 / 0)
Why shouldn't we expect the Democratic nominee to be trouncing his opponent right about now? This isn't an "emotionally based" reaction - it's a rational assessment; this is one of the most favorable years for Democrats in quite some time, yet Obama has a minuscule lead.  
[ Parent ]
many reasons why (0.00 / 0)
Obama can't singlehandledly overcome decades of negative branding re: the Dem party, as well as the distrust of gov't that the GOP has sowed systematically during this multi year effort.

We need much more than just Obama; his campaign is a great start, not the solution.  Even in 2006, D's did not wipe the floor with huge margins.  But "shouldn't" Democrats have crushed R's everywhere that year?  

The country is not overwhelmingly capital-D Democratic, even tho the country may in the aggregate hold more liberal positions than conservative ones.

And "trouncings" are historically very rare.  

[ Parent ]
Huh. (0.00 / 0)
So your theory is that McCain hasn't picked up any new voters, he's just been peeling off undecideds who had been inclined to vote Republican but hadn't connected with McCain to the point of considering themselves supporters until now, when they started hearing the right dog whistles? That's an interesting idea-- do you think that if it's true there's some kind of polling that could be performed or looked at to nail its accuracy down for certain?

This theory also seems to raise the question-- the remaining 9% or so "undecided" according to Pollster. Who are they? Are they mushy middle? Are they more Republican-leaners who McCain needs to just keep pandering to more, or who Obama needs to convert into Obamacons? Are they Democrats who are sitting in the undecided column because Obama isn't pandering to the left enough? Is there a way to find out who these people are?

"Should be up 10" (0.00 / 0)
I don't argue with polls, which currently state a close race in the popular vote and a substantial wide-but-shallow lead in the electoral college.

But those expecting Obama to be up more in the polls, ignore the fact that in many election, particularly those involving less known, less defined challengers, the race remains close until just before the end.  While we don't know if the election will be close or not, if we assume an Obama landslide will happen, for the sake of argument, there is no reason to expect to see a big lead this early.

In other words, yes we can be concerned about the day-to-day.  We can be concerned that Obama might get defined in unfavorable ways and the be unable to redefine himself.  But it does not follow that a close race now means a close race in Nov.

While not a reason to put on blinders or get complacent, we should bear in mind that the environment is so favorable to Democrats that a McCain victory would basically represent one of the biggest upsets in American history.  Which is not to say it can't happen, but I think the pessimists gloss over or hand-wave that away far too much.

Back To A National Toss-up | 22 comments
USER MENU

Open Left Campaigns

SEARCH

   

Advanced Search

QUICK HITS
STATE BLOGS

Home | RSS/XML Feed | Register | Contact

  © OpenLeft 2007-2010, Powered by SoapBlox

Powered by: SoapBlox