November 2, 2012 Analysis
Polls included:
Tracking Polls
R+2 - Rasmussen Daily Track
R+5 - Gallup
O+1.3 - IBD/Tipp
O+0.07 - ABC/WaPo
Others
O+1 - Battleground
Even - Pew
R+1 - NPR
O+1 - CBS/NYT
Even - Fox
O+5 - National Journal
O+0.04% - Current RCP Average
O+0.61% - Average using the 2008 turnout model
R+2.18% - Average using the D+3 turnout model
R+4.42% - Average using the 2010 turnout model
R+4.47% - Average using the 2004 turnout model
R+6.20% - Average using the Rasmussen Party ID turnout model
Two things happened yesterday in the RCP average. First the Rasmussen didn't change even though all of the weekend polling dropped off. The last two days of polling matched the Sat/Sun numbers, which generally means a shift toward Obama. However, keep in mind that Romney is still leading by 2 in a D+3 poll. This is a good lead. It is also possible that the hurricane is disrupting their polling operations since they are located in New Jersey. Additionally, Rasmussen has switched to 1000 samples every day, from their typical 500 samples. I don't know how it could be affecting the results, but I wanted to put it out there.
The ABC/WaPo tracker also shifted very very slightly toward Obama. I am listing the actual lead above, instead of the +1 that RCP is listing. The sample moved to D+4 from D+5, so it is a legitimate shift in the poll. This is mostly due to Independents moving from favoring Romney by 6 to 3. One more point about this poll is that 20% of the respondents say they have already voted, while 19% say they plan to. I find it difficult to believe that 25 million people will vote over the weekend, especially since many large states, like Pennsylvania, do not have in person voting.
So after discussing all the movement toward Obama, let's keep in mind the real state of the race. Even with the terrible National Journal poll in the averages, Romney is still leading by over 2 points with a D+3 turnout. And with the 2010 turnout that I personally expect, Romney will win by over 4 points. If we drop that National Journal poll out of the averages, Romney leads in these two models by 2.4% and 4.6% respectively.
This is a bit like in 2008 when McCain had a minor surge in the last week, but it wasn't enough to overcome Obama's lead.
DIF,
ReplyDeleteRe: Rasmussan's shift to a larger sample size.
I don't know where I heard this, but I saw somewhere (man I hope it wasn't here) that polls with a larger sample size tend to favor Romney more.
There are a number of possible explainations for this (especially if there's a percentage of people skipping through the LV screens who are actually not at all LV's and just randomly picking a canidate.
I must say i am worried now.. Rassmussen moved to a 48-48 tie from 50-46 romney lead a week ago. What is going on? This is making me nervous.
ReplyDeleteWhere are you seeing that? As of 10:06, Rasmussen hasn't reported the tracking poll for Friday.
DeleteNow I see it. Maybe the hurricane was a boost for 0bama after all. Not sure that will be decisive when people go vote on Tuesday, but we'll see.
DeleteIn addition to my reply below, I will be interested to see what Ras (and Gallup) says on Monday. Of course, it might not matter and Obama has already lost this given the gap between Rep and Dem enthusiasm.
DeleteIn any case, come Wednesday there could be a lot of explaining to do...
Surely appears that there post-Sandy, that Obama is the leader at least the last couple of days. Poor Romney, ran a good campaign, debated well, but might lose because of something entirely outside of his control.
ReplyDeleteRomney wins R's by a wider margin than Obama does D's and wins independents by 3 still in the new Ras? Did I read all that right? That almost has to mean that Ras has gone back to a bigger Dem sample, no? Those numbers don't add up otherwise? What in the world did he do that for?
ReplyDeleteMy sense is there's a Sandy bounce, but temporary. Should tick back down. Obama may net a .5 - 1 pt gain out of the Christie bi-partisan love fest.
ReplyDeleteI tend to agree. I understand what Christie was doing but still pisses me off because all that was was a photo op for the both of them. Between that and his speech at the RNC, I have to wonder if there is no love lost between Christie and Romney.
DeleteAnyway, I still think nationally Obama is still in trouble and that none of these polls have a clue because it is such an odd election. If Romney is to win, hopefully the reputable ones make the necessary changes because the MSM seem to take all of them at face value.
Temporary is all he needs, the election is Tuesday. I can't believe the timing on that storm worked the way it did. Maybe 0bama really does have power over nature.
DeleteThe Sandy thing could also tick down for the prez if we start getting stories about red tape and squalid conditions.
DeleteBut I agree, Christie could have been a lot more professional in his Obama interaction. One thing to work together, the other thing to give a bunch of emotional praise.
Christie acted like someone ticked off that he wasn't the VP pick. He may be thinking about running on 2016, but if he thinks this helped his chances at being the GOP nominee some day, I think he blew that calculation, and I don't think he had much chance at ever being the nominee in the first place.
DeleteThat Ras number is concerning since they're one of the few pollsters I actually trust (somewhat).
ReplyDeleteRasmussen confuses me. In 2008 his party identification was pretty close to the party composition via exit polls (D+7). Today, the latest party identification figures he has shows R+2.6, yet is basing his model on a D+3 sample. What am I missing?
ReplyDeleteAnyway, the Rasmussen numbers are concerning. It's going to piss me off to no end if Sandy/Christie end up giving Obama the election.
Ras has it tied and I am seriously about to have a heart attack over this whole thing. Rose colored glasses off, it looks like Bambi is actually taking this. I don't even know how it's possible. I pray I'm wrong.
ReplyDeleteexactly.. mega independent. I was sleeping well till last night and now this.
DeleteWell, I hope Dave can dig up the internals on this one and show that R is still leading the indies..otherwise, this will be one anxious weekend for me :)
Based on what? Ras is Zero's best case turnout, and at 48-48 the incumbent still loses.
DeleteWe gotta see what 48-48 becomes by Monday. I'm back to being confident having looked at the party registration numbers in swing states (and common sense), but that number almost killed me.
DeleteYou are wrong.
DeleteRomney is still leading in Ras by .8. It's a rounding error. And Ras is still sampling D+3.
Whew. That sounds a bit better at least.
DeleteMega - I'd be curious to see what you dug up regarding party registration.
DeleteNot to worry, it's a sandy bounce that will dissipate by Tuesday. Could actually go the other way, have you seen the Marathon story? Just incredible that they are running it Sunday amid the circumstances.
ReplyDeleteThat and some photos of poor conditions could make this a negative story for Obama "It's the same as Katrina"
Polling numbers are like numbers reported by the BLS. You don't know for sure what the real numbers should be; you're just pretty sure that what they are reporting are *not* the real numbers. Ras has been acting kinda odd this whole cycle. (shrug) Best thing to do at this point is donate or volunteer if you can, and make damn sure you go vote against Barky. Not much else we can do now.
ReplyDelete@ roy...I'd agree if I thought anyone in the MSM would actually report it that way. Maybe word will get out anyway...it seems to have about Benghazi, at least to some degree.
ReplyDeleteNormally true, but the media can't resist these stories, natural disasters are their absolute favorite.
ReplyDeleteI wouldn't hit the panic button just yet, I see this as a "stopped the bleeding" moment that has already dissipated. This was a hurricane that hit a few blue states, it was not 9/11, it wasn't even Katrina. And the aftermath could even give the Administration a black eye for mishandling.
ReplyDeleteAre Americans really going to give Obama 4 more years over a photo op with Chris Christie? I don't think so. If Obama wins Tuesday, he was already going to win anyway.
If the electorate looks like 2008, it was over before it began. If the electorate is closer to what it looked like in 2010 or 2004, it won't even be close, Romney will easily win.
I wonder what Ras's electoral snapshot for October is.
ReplyDeleteWhat would a D+2 turnout look like?
ReplyDeleteRomney by about 1.5%
DeleteJay Cost predicts Romney win.
ReplyDeleteNothing in the analysis we don't think ourselves, but it's nice to have 2004's guru in the fold.
I guess I need to post something about the Rasmussen update, since everyone is panicking.
ReplyDeleteI spent all morning adding up the Ohio actual vote numbers. Silly me.
LOL. Poor Dave...forever having to sooth the nerves of us Nervous Nellies. Thanks for all you do here. It is much appreciated even if the effects do seem to wear off of us faster than you'd like. :-)
Delete