1. So Romney is still up by .8% in the RCP head-to-head, but Obama’s average job approval has rebounded to 49.7% from its weekend low of around 48%. So we’re back to wondering if the remaining undecideds will break in the direction of the challenger or the direction of the incumbent’s job approval.
2. Romney and allies are making a heavy media play in the Southern New Hampshire media market. Crossroads is doing most of the negative media. Romney is doing positive ads whose theme can be best summed up as “cooperation for solving problems through bipartisanship like I did in Massachusetts” The ad is well produced, but it doesn’t really work for me. Then again, I’m not an undecided married white woman so I’m not the target audience. I can’t count the number of Romney ads I’ve seen lately compared to exactly zero Obama/anti-Romney ads.
3. I can’t remember a year where the polls have been this close, this late, for this long (Bush-Gore came down to a tie in the last weekend.) Am I wrong to find it incredibly boring?
October 29th, 2012 | 5:11 pm
Maybe if Rodney Dangerfield as Al Czervik were on the ticket …
October 29th, 2012 | 5:16 pm
You don’t live in Ohio.
October 29th, 2012 | 5:33 pm
Pseudoplotinus, he sort of is (minus the wit, private sector accomplishments, good taste and self-control.) Paul Ryan handled it a lot better than Judge Smails.
October 29th, 2012 | 6:16 pm
O’s been sitting at 47 for a while now. How does an incumbent make it from there to 50 at this point?
October 29th, 2012 | 7:04 pm
Brian,
Last Thursday over at AEI there was a panel discussion on the state of the election and Henry Olsen and Michael Barone had a good exchange about the 47% issue. The discussion begins at 40:30 regarding where the undecided vote will go, and Barone interjects at 42:00 regarding the 47% point for the president and its meaning.
http://www.aei.org/events/2012/10/25/aei-election-watch-2012-session-6-down-to-the-wire/
Basically Olsen is pessimistic and Barone is optimistic for Romney.
October 29th, 2012 | 7:13 pm
Brian, at this point, George W. Bush was at 48.1% (Obama is at 46.8%) in the RCP average of head-to-head polls and he went on to win by 2.4%. So Obama is on track for a tiny popular vote plurality? But wait…Romney is also running 1.6% ahead of John Kerry at the same stage even as Obama runs 1.3% behind Bush. So we’re on track for an even more microscopic Romney popular vote plurality right? But Bush’s job approval rating was 48.9 (or just a shade lower than Obama’s – to the extent that differences of a point are even meaningful at all.) But late deciders broke 60%-40% for Kerry in 2004.
So it is all just a mess of closeness confusion.
October 29th, 2012 | 7:37 pm
My bad. Bush was at 48.8 today in the head-to-head. He went down to 48.1 four years ago tomorrow.
October 29th, 2012 | 7:48 pm
Pete: Hmm. I know I’ve seen Gallup (scroll down a bit here) tracking numbers for Bush from 2004 showing him consistently sitting at ~50% or so throughout October. O’s not anywhere near that high in the polls anymore.
To me the thing making it clear R’s winning right now is the outright lunacy of the O campaign these last few weeks. We’ll know all in just over a week…
October 29th, 2012 | 7:54 pm
I don’t want to rain on anyone’s parade, but there are two quite distinct and different questions being asked here: If the election were held today, who would you vote for? and, Do you approve/disapprove of President Obama’s job performance?
There is no way on earth to defend these two as logically symmetrical to each other since there are four possible choices involved: Approve/Romney, Disapprove/Romney, Approve/Obama, and Disapprove/Obama. All of these choices are possible and the last three of them are virtually certain to have a significant number of respondents.
Individual undecideds logically cannot break in the direction of either approving or disapproving of Obama’s job performance since this question is not on anyone’s ballot. Moreover there is no evidence what that those who actually answered these either\or questions were undecided about anything. How you answer the question of who you will vote for with your made-up mind has no connection to how I will answer it from my undecided mind–or if I will even answer it at all.
The only logically legitimate way to relate the answers on one question to answers on the other is correlative, and not as an either/or choice. At least as far as this post is concerned, the proper numbers are not present to do that.
You can correlate the vote Obama score to the approve Obama score, but you cannot correlate the vote Romney score to the approve Obama score since the question doesn’t in the least ask anything about Romney. And you can’t either compare or correlate “the percentage Romney is ahead” to any of it.
In other words, the reason you are bored with this data might just be that you are not analyzing it in a way that can possibly lead to any significant conclusion.
October 29th, 2012 | 10:22 pm
Joseph, rain any time (on me anyway), but given that about 5-6 of the likely voter population remains undecided I think it matters a lot in how they break based on what proportion are Obama-approvers or non-approvers (I think that all but the most idiosyncratic DISapprovers are in Romney’s camp already.) Or two put it this way, if they were overwhelmingly Obama non-approvers, the way would be open for a mini version of 1980 where the undecideds and soft incumbent supporters stampede to the challenger. That doesn’t seem to be the case now.
Now to your comments on whether Obama non-approvers will go to Romney or not. I think most of them will. There are of course some non-approvers of the “I disapprove of Obama because I really want Bernie Sanders, but voting for Ralph Nader worked out badly” variety, but I’m not worried about them. Romney is just good enough of a candidate that I don’t expect much leakage of the remaining Obama non-approvers to the incumbent (he isn’t a Sharron Angle), but he is not nearly so good a candidate that I expect people who think (even mildly) that Obama is doing a good job to vote Republican for president in large numbers.
So you have this situation where, absent a decisive event, something has to give.
October 30th, 2012 | 9:08 am
We really don’t know–no polling data is meant to precise enough to give us confidence one way or another. So I fall back on the fact that Romney has to carry more toss-up states than Obama to win. If I had to bet the farm to save the family on this, I’d bet Obama. But I’d rather not bet on something that I really don’t know enough to think I have any kind of insider advantage.
October 30th, 2012 | 11:04 am
From the nerd point of view, you have to step beyond that and look into why we don’t know. First, there is random variability. Second there are “house effects” based on individual pollsters’ criteria to both select and evaluate the polled. Third, there are exterior considerations that may be baked into the data.
Random variability can be largely reduced by mathematical tools to give % of likelihood. PEC touts the specific choice of using the median rather than the mean to establish the data’s central tendency because it filters out house effects and other outlier problems. Exterior considerations can be minimized by careful choice of which polls you evaluate and how consistently you evaluate them (the typical choice is State over National polls and vice versa).
Because we can define where the uncertainty is coming from, we can adapt to it and reduce it. The fundamental principle is that people in crowds behave in predictable ways even though there is considerable uncertainty about individual behavior.
Much of the suspicion and hostility toward polls, and the nerds who work with them is that the principle of human predictability offends our amore propre.
But, no matter what kind of personal huff we get into about it, it is still true. And if it weren’t, polls would have no point whatever.
Links
Blogs
Find Us
Contact