TNR BLOGS

July 05, 2009 | 4:05 PM
July 05, 2009 | 12:13 PM
July 04, 2009 | 11:18 PM

March 09, 2009 | 5:19 PM
March 09, 2009 | 5:16 PM
January 07, 2009 | 12:20 PM

July 05, 2009 | 12:02 PM
July 01, 2009 | 10:33 PM
June 30, 2009 | 8:42 AM

July 26, 2008 | 2:24 PM
July 23, 2008 | 1:55 PM
July 17, 2008 | 3:56 PM

July 03, 2009 | 10:13 PM
July 02, 2009 | 12:57 PM
July 01, 2009 | 7:02 PM
COLUMNISTS
TODAY'S STORIES
02.11.2008
Today's Polls: Some Tighetning Toward McCain in State Polls

Polls conducted since our update last evening suggest some tightening toward John McCain, but he sits well behind both nationwide and in many key battleground states and remains a long-shot to win the election.

The good news for McCain? SurveyUSA has become the latest pollster to show the race tightening in Pennslyvania, now giving Barack Obama a 7-point lead after he'd been in the mid-double digits at various points in October. The Muhlenberg/Morning Call tracker has also continued to tighten, also settling on that 7-point number.

SurveyUSA also has Virginia tightening a bit to 4 points. And McCain gained incrementally in the Research 2000, Gallup, and Diageo/Hotline trackers, although this comes after a couple of days when Obama had been moving up. (Rasmussen held steady, whereas Obama ticked up in Zogby).

Overall, our model shows McCain closing Obama's gap in the national popular vote to about 5.4 points. His win percentage has increased to 6.3 percent, from 3.8 percent last night.

However, several cautions about reading too much into these numbers:

Firstly, I have the model programmed to be EXTREMELY aggressive this time of year. There have been relatively few 'fresh' polls conducted within the past 24-48 hours -- most of these state polls were in the field late last week. As we get more data in today and tonight, the model could very well decide that the race is not tightening at all. Moreover, polls conducted on a weekend -- particularly on a quasi- holiday weekend -- is generally unreliable.

Secondly, even with this tightening, McCain remains well short the 2/2/2 condition that we defined last week:

John McCain polling within 2 points in 2 or more non-partisan polls (sorry, Strategic Vision) in at least 2 out of the 3 following states: Colorado, Virginia, Pennsylvania.
Indeed, McCain has not come within 2 points of Obama in any polls in any of these states.

Finally, where McCain has made progress, it has come mostly from undecided voters rather than Obama's support -- this is particularly the case in Pennsylvania. Therefore, he may be running out of persuadables to persuade.

There's More...

--Nate Silver

Posted: Sunday, November 02, 2008 2:02 PM with 12 comment(s)

Comments

You must be logged-in to comment.

Not a subscriber? Click here to get a digital or print and digital subscription to The New Republic!

Mack2 said:

Research 2000/DKos poll reported today says Obama has 7 point lead, not 4. I assume you'lll repair this for your evening summary.

November 2, 2008 2:52 PM

jet said:

Speaking of polls and voting.  No one I've talked to, and that's several, is voting Tuesday.

All have already voted, many quite a while ago.  They'd seen what they needed to see is what they've said.

I didn't ask them who they voted for, but many cited the financial crisis and how the candidates handled it as sealing the deal.  Debate performances were also mentioned.

November 2, 2008 3:05 PM

propositionjoe said:

All of these polls indicate to me that McCain has no chance of winning the popular vote. Obama is going to roll up huge numbers in high population states and if he loses enough battleground states to tip the  election towards McCain, the losses will probably be narrow. I hope that the national numbers suggest that a wave election is likely to take place so that electoral numbers match the big popular vote margin that he is likely to produce (Knock on wood). But all of this information suggests that someone (any leaders out there?) needs to try to fix our 3rd World electoral system. Ten hour waits during early voting sessions? Really? How is that necessary or acceptable? A clear majority's will denied due to a rickety electoral system designed when there were fewer Americans in the whole United States than there are in contemporary New York City or LA? It is absurd. And Democrats don't have to be on the losing end of an election to suggest that our system is inane and unfair. REFORM IT.

November 2, 2008 4:07 PM

cspencef said:

"There's more"

Is that supposed to be a link?  It looks like a link, but it doesn't link...

November 2, 2008 4:09 PM

propositionjoe said:

I checked FiveThirty-eight.com, and the "there's more" is a phantom. The totality of the post has been presented.

I need someone to tear the internet from my cold, anxiety-ridden fingers.

November 2, 2008 4:33 PM

JEFF FREY said:

The Gallup polls of the last few days have an interesting feature. Obama's margin over McCain been almost identical whether the version used is for registered voters, or either of their two likely voter models. As recently as a week ago, their expanded likely voter model gave a tighter margin than registered voters, and the traditional model tighter still. Details are at Gallup's website:

www.gallup.com/.../Gallup-Daily-Obama-Continues-Outpace-McCain.aspx

So why the change in this aspect of the poll results? I hope Nate weighs in on this, but here are my own thoughts. The expanded model includes people who have already voted as voters (duh), regardless of other factors. Widespread early voting might explain how this model would approach that of registered voters, but it seems that the traditional model does not factor in whether or not people have already voted (except that, presumably, those people would have the maximum score on the do you intend to vote question). So unless Gallup is now including early voters as automatically in the traditional voters model, which they don't say they are, I don't see why the gap in that model would have expanded to match the registered voters model. A non-representative sample does not seem to be that liklely either, given that this has persisted for a week or so. Another possible explanation is that McCain voters are letting their lack of enthusiasm show, because Obama gets the same fraction (51-52%) in all three models.

Whatever the cause, attempting to analyze the polls beats fretting about them. Especially because I was traveling to a swing state last week (VA), and got to see the whole suite of McCain attack ads (not really that good) and Republican 527 group ads (uniformly nasty, based on gross distortions, and including plenty of code words to appeal to the lowest impulses).

November 2, 2008 5:22 PM

jacobt1 said:

JEFF FREY said,

"and Republican 527 group ads (uniformly nasty, based on gross distortions, and including plenty of code words to appeal to the lowest impulses)."

Did you see any of Obama uniformly nasty, based on gross distortions, and including plenty of code words to appeal to the lowest impulses adds, including ads that telling seniors that McCain would take their SS and Medicare away?

November 2, 2008 5:56 PM

jacobt1 said:

" But all of this information suggests that someone (any leaders out there?) needs to try to fix our 3rd World electoral system"

Let's start with making sure that people who donate to campaigns don't break the rules, tlet's create a system that would not allow people like Zeituni Onyango to donate or vote.

November 2, 2008 6:02 PM

ironyroad said:

There could well be legitimate reforms needed to tighten up the small-$$ fundraising process, but there is not a shred of evidence that Onyango or anyone else has voted who isn't legally entitled to, and even if there is a tiny percentage, those people are more than outweighed by the large numbers of people who are entitled to vote but cannot, due to bureaucratic screw-ups, confusing application of regs, inconsistent status of provisional ballots, and outright suppression.

In recent days, Rhubarbs and other posters on this board have analyzed, with numbers, the impossibility of any kind of voter ID fraud.

Let's keep things in perspective.  We already know that the Republicans will go as far as firing U.S. Attorneys because they do their constittutional duty and refuse to bring prosecutions for which there is no evidence.

November 2, 2008 6:18 PM

JEFF FREY said:

You need to try harder than that, Jacob.

November 2, 2008 6:23 PM

jacobt1 said:

Jeff, It would make no difference for you. The One is always right and pure.

November 2, 2008 7:53 PM

JEFF FREY said:

Wrong again, Jacob. You have no idea at all. What I mean is that if you want me to take you seriously, you  have to write posts that are thoughtful and intelligent. A good start would be to stop the cut and paste, and stop repeating people's arguments with a noun or two transposed. Do that and I'll be happy to argue with you. Continue as you have, and you are just being a troll, and not worth the time to argue with.

November 2, 2008 8:16 PM