The biggest victim of our post-truth 2012 political campaign has to be the field of statistics:
In the two new national polls, President Obama and Romney are now tied among all registered voters. In Gallup, this is change from a five-point Obama edge in the three days leading up to the Denver debate; for Pew, it is a shift from a nine-point advantage for Obama in mid-September.
So who moved in Romney’s direction?
Well, not political independents, for one. There was no meaningful change in their support for Obama or Romney in either poll.
All of the change in both polls came from the composition of each sample. In pre-debate interviews by Gallup, self-identified Democrats outnumbered Republicans by five percentage points, according to Gallup’s Jeff Jones. By contrast, in the three days following the debate, the balance shifted in a GOP direction, with 34 percent of registered voters identifying as Republicans (two points up from pre-debate), 33 percent as Democrats (four points down).
For Pew, a nine-point Democratic advantage in mid-September is now plus one percentage point for the GOP. (The turnabout in “likely voters” was even more dramatic, shifting from Democrats up 10 to Republicans up five.)
It’s all part of the GOP plan to erode public trust in anything, well, public.
AT
But Sully told me that it’s over, Obama is purposely trying to lose!
Eric U.
how did you sneak this post in after the liveblog one?
piratedan
just promise us that you’ll never change, you big lug… after all, we’ll always have Tunch and they’ll never take that from us
freelancer
Sully’s running around NYC with his hair on fire saying Obama’s throwing the election on purpose.
ETA: AT beat me to it.
Alison
JC, got a link for this? I’d like to share it with a couple people I know who are going all mouth-foamy with worry…
dollared
So how is that a problem? Republican leaners are emboldened because Obama acted like a loser in public, fitting entirely with the narrative that $1B in ads have been portraying – that he’s not up to the job.
So now they happily say they are Republican, because they like to be associated with a winner.
How are we going to turn it around? Because it is looking like it will need a turnaround.
Matt McIrvin
This isn’t, in and of itself, a sign that something’s screwy with the Pew poll. Party ID is fluid, and for many people it’s basically a proxy for who they support in the presidential race. We shouldn’t dismiss results because of the party ID mix in the sample any more than the Unskewed Polls people should.
That said, this result seems out of line even with other polls during the same time period. I suspect part of what’s going on is that Romney was leading the race for a short time Thursday and Friday, so results end up being very time-sensitive.
eemom
Dear John Cole, I have it on good authority that if you didn’t suck and would buy more bandwidth or something then poor DougJ and mistermix wouldn’t have to do whatever that thing was they did on the last thread. I’m sure if you rattled the jar everyone including me would pitch in and all the drunks in the world could hit refresh on election night and STILL the site wouldn’t crash and we’d all have us a kick ass partay and everything would be groovy. love and kisses, eemom
Shawn in ShowMe
@dollared:
I thought all this angst was supposed to be a result of Obama “losing” the debate? I assume Obama “winning” a debate would turn it around. Or do these polls only swing one way?
OC
@AT: Have you read his last post today? Good grief the man is a hysterical mess.
Valdivia
so wait–a week ago the number of dems was off the charts and like in 2008 and now we have gone back to 2010 levels? wut?
AliceBlue
I mentioned it on an earlier thread, but anyone who’s still freaking out needs to read Nate Silver’s latest post.
burnspbesq
I find the idea of a “battle for the soul of the Republican Party” to be a least a little absurd, but if this is where the battle lines are going to be drawn, make more popcorn.
http://www.washingtonpost.com/politics/decision2012/in-missouri-clergy-in-the-fray-of-akin-race-seeing-it-as-start-of-a-battle-for-the-soul-of-gop/2012/10/08/500ab2c4-0eff-11e2-bd1a-b868e65d57eb_story.html?wprss=rss_campaigns&tid=pp_widget
piratedan
well judging from the cacophony regarding one small advancement of the Mittster not looking like a complete fool when given the opportunity in comparison of a campaign filled with gaffes, you didn’t really think the media would play fair when given a glimmer of hope for their narrative didya?
All we found out is that there are still a number of folks who are holding out hope that this guy (Romney) still isn’t a complete whack job; thereby allowing them to cowardly vote for the Vulture Mormon and assuage their white guilt while they flush the dreams of the country down the tubes telling themselves that there’s no way he could be THAT BAD, really, honest, truly, maybe, I sure hope not…..
Hill Dweller
@Matt McIrvin: As we talked about on the lower Gallup thread, the Pew poll sampling is very shady.
Their registered voter screen started at R+1, which is damn near impossible. Non-white voters make up just 15% of the likely voter sample, despite the projected percentage in the election being 28. Voters 50 and over made up 69% of their sample. Voters 18-49 made up just 30% of their sample, despite being 57% of the electorate in ’08. Breakdown of voter geography slanted South: Northeast 201, Midwest 271, South 417, West 223.
I could go on, but you get the picture. It’s an embarrassing poll from Pew, which has always been credible.
Alison
(Thanks for adding the link!)
I swear, I’m more irritated with freaked-out Dems than I am with Romney at this point. Or almost more…
burnspbesq
I will admit to having freaked out a little when I got an email from MoveOn alleging that Sherrod Brown is trailing.
Tractarian
Sullivan has absolutely gone off the deep end. The stuff he has been posting lately is really just indefensible. (He apparently could use a few more mental health breaks, I’m guessing.)
Don’t get me wrong, I appreciate Sully’s Obama hagiographies as much as the next guy. But for a ‘seasoned’ political observer, he is remarkably naive about polling. He takes one poll, treats it as the final word, and then picks it apart to show all the different ways in which it reinforces his narrative. He doesn’t seem to understand that polls are random samples, and while the fact that a poll is strong for a candidate can be important, it really does not help to point out that the poll’s internals are also strong for that candidate. He did that a few times with Obama, now he’s doing it with the Pew poll showing Romney ahead.
He did not seem to realize that a slew of swing state polls came out today showing Obama ahead (including some Rasmussen polls), and that all the polls showing poor results for Obama including at least some samples from immediately after the debate. The situation is nowhere near as dire as he describes.
John Cole
@eemom: You are wrong.
Sure, I could rattle the tip jar, and get donations, but it makes absolutely no sense to be running a website that gets 75k hits a day on a server that could handle 250k hits a day.
The money is not worth it. I assume you all have families and bills, so I try not to use the tip jar unless something really requires it.
But back to the point, the issue is server performance. The server works perfectly 99% of the time. Every now and then, we have a debate or an election, and shit craps up. Why pay to have the capacity for three days a year, when what we have works 99% of the time.
The coveritlive plan is what is known as an elegant solution. We will not require massive server upgrades, we will not require massive wasted donations from readers, but we will be able to cover the debates and the election.
I know this will not make you happy, and I know you are actually not concerned with server performance- I bet you couldn’t give a shit. You just like complaining. The only thing that makes you happy is being miserable and shitting all over me, but, you know, hey- I like the solution DougJ and Mistermix came up with, even if miserable fucks like you are… miserable still. Because I know the only reason you read this website is to be an asshole to me.
Oh, and to be a pedant.
mai naem
Okay, I am going to admit I’m worried now. I get the statistics bit. I never trusted Kos’ polls and I’ve known for a long time that Rasmussen was a Repub. poller. But, jeebus, there’s a Michigan poll with Romney within three points and a Repub Penn poll with Romney within a couple of points. WTF? Obama was bad at the debate but OMG people are going to base their votes on the debate????
On the other hand, I am so freaking sick and tired of this crap, if this freaking country wants Romney/Lyin’ they freaking deserve all the destruction they’ll bring.
And, yeah, I go to 538 every freaking day just to calm the fvck down
Not Sure
@Hill Dweller: No way us old folks make up nearly 70% of the electorate. We’re strong, but we’re not that strong.
/OK, I’m 49.
//Close enough for government work.
TexasMango
@dollared:
The problem is there actually are more registered Democrats then Republicans. If there are suddenly more Republicans then Democrats and Independents are split winning this elections was never going to happen anyway.
There are all kinds of things about this poll that aren’t reflective of the electorate. Here a re just a few:
Jeff Gauvin @JeffersonObama
& 70% over 50 RT @pollbuster: Pew should truly be embarrassed to release a poll with #s : Northeast 201 Midwest 271 South 417 West 223,
You think the electorate is going to be 70% over 50, 83.4% white and 38% Southern? Talk about skewed polls.
scott
The motivations of Rasmussen and We Ask America are pretty explicitly partisan. I really don’t understand the Pew result on those grounds, though. Sometimes party affiliation results in polls can shift because events change the way people think about themselves, so maybe that explains it here (the debate). Still, that seems a pretty radical shift and doesn’t explain some contrary indications in Gallup, even Rasmussen, and some of the state polling the last couple of days. I suspect we’ll need more polling to really sort it out – I agree with Matt McIrvin that it may be a function of a pro-Romney spurt over a couple of days.
For whatever it’s worth, in the (I know) kinda offbeat but still interesting Rand Poll, over the last 3 days you had a 1.5 point shift to Romney, a 0.6 shift, and then a 0.1 today. The other continuous trackers, Ras and Gallup, also seem to be indicating a lessening bounce. Stay tuned, and hope Biden gets on the board.
scott
@John Cole: If most of the places I had to commute to were within a radius of three blocks, I wouldn’t buy a 100,000 dollar gold-plated Hummer with the special surface-to-surface missile option either. Makes sense to me, especially since it would be your money or (even worse lol) ours.
Hill Dweller
@scott: Pew had a registered voter screen of R+1. That is horseshit.
Gin & Tonic
Shorter Cole: Capacity planning, how does that work?
muddy
The “self-identified” part is stupid. I always tell them I’m a Republican that prefers Obama.
I got one this evening and they started out by saying they only wanted under 45 or over 60. I said I was in between. Afterwards I wished I’d lied so I could pretend to be 38 and love Obama. They caught me off guard with the age thing.
I answer these things because if I get a push poll I notify the campaign. The push polls have always been local and they are happy to hear about it. And they are always but always against the Democrat.
I have also gotten 2 things ready printed with all my info to ask for an absentee ballot, both from a conservative PAC or wev it is. They don’t say they are conservative, but the prose is very clear. My friend told me that her husband got one and she didn’t, I wonder how they pick and choose? I shredded both of the ones I got, supposing that they are shitty enough to sift for the paper at the transfer station.
I like to vote in person, it’s like when we used to have town dumps and Saturday morning was a big social event. The senior citizens group has bake sales, and the girl scouts too. Calories don’t count that day because they are burned up with the fire of patriotism.
Or something like that.Suffern Ace
@scott: Gallup will be releasing a new tracking poll tomorrow that shows the candidates tied. That’s the seven day poll. Before we go nuts and walk down the polls have a conservative bias road, it is possible that what is being measured is real. I will say, this election was going to be close. I expect that Obama will win, but we’re talking 2000 levels of closeness.
AT
@Tractarian:
Day to a “sorry I was wrong should have known Obama plays a long game, meep meep” post is now T-4?
Cam
If Romney really got a bounce from his overrated debate performance, it seems pollsters wouldn’t have to slant things as hard as it seems PEW did, per Gauvin and others who looked at the crosstabs.
Comments from Jeff Gauvin @Jefferson Obama and @pollbuster on the PEW poll :
“PEW increased GOP weighting +5% & decreased Dem weighting 5% & VLAM….instant Romney Surge narrative is born & fought over on TWITTER #POLL”
“PEW POLL just like CBS flash debate poll…mostly over 50% (70%), Southerner and few Latnos & blacks. Brutal cross tabs”
http://twitter.com/JeffersonObama/status/255469589176139777
“70% over 50 RT @pollbuster:
“Pew should truly be embarrassed to release a poll with #s : Northeast 201 Midwest 271 South 417 West 223”
http://twitter.com/JeffersonObama/status/255469324167426048
“PEW POLLS —–15% Non White vs 28% Expected to vote in 2012—-NO LATINOS”
http://twitter.com/JeffersonObama/status/255459802191384576/photo/1
Richard Fox
I remember reading a story about one of the writers of the Twilight Zone, if I recall correctly his name was Buck Houghton. He would go out with his pals, and would take normal occurrences– a closed door, or stray noise, say, and whip up such hysteria over what it might have been, who might be lurking under there, etc, that after a while he had them all scared out of their wits. Basically they starred in their own miniature Twilight Zone episode. I look at the last several days and I think of that same quality of hallucination. If it wasn’t the debate or the polls I am sure it would have been something else to get folks unhinged, life will end, etc. The one saving grace in all this noise is 1) this IS Mitt Romney folks are worried over, the worst candidate ever in the history of the world, and 2) the media loves a comeback story. So the elements of Romney screwing up something and Obama hitting a debate epic win are all there, waiting.. growing… starting to coalesce in the muck.. And voila! A new meme is born! :-) Have a good night, all.
TexasMango
@Or something like that.Suffern Ace:
I always thought the popular vote would be close because R-Money is going to win 90% of the white vote in the Deep South. I don’t believe the EC will be close.
Mark S.
@Hill Dweller:
What does that mean?
nepat
@scott:
The Pew poll is really just the result of the right-wing going apeshit for three weeks about skewed polls. How did pollsters handle that pressure? They buckled and “unskewed” them. This is why Repubs are constantly bawling like infants about every little thing. It pays off. Notice how no one’s talking about the favorable-to-Obama jobs report? Not allowed. The right wing said it’s a ginned up fake.
Shawn in ShowMe
@Or something like that.Suffern Ace:
I’m still confused. Why does polling taken after one debate performance in early October indicate an election that will be 2000 levels of closeness on November 6?
Hill Dweller
In that Pew poll, Obama went from leading by 3 with white women in their mid-Sept. poll to losing by 19 in this poll.
Pew’s poll sample is laughably white, male, old, southern and Republican. Willard leading by just 4 in this ridiculous poll should actually concern his campaign.
Matt McIrvin
@scott: The Rand poll is a weird one, in that nothing happened there as a result of the debate. Nothing, or so little as to be indistinguishable from nothing. There had been a little movement toward Romney in the days before the debate, nothing a Democrat could really get worried about. It’s continued since then. But the rate didn’t really change in any significant way through Wednesday.
For the first couple of days I actually tried to explain the lack of movement by saying the data was somehow delayed by a day and couldn’t possibly be post-debate. But it’s continued that way.
It could just be that the Rand sample was of people who, just by virtue of being in the Rand sample, were unusually plugged into politics and therefore unlikely to be swayed by one TV debate.
James E. Powell
@John Cole:
I like this coveritlive thing. It should be interesting and amusing.
Of course, if you really want it to be entertaining, you should invite Sullivan to participate.
Or something like that.Suffern Ace
@Tractarian: I do agree. The whole Obama is throwing the election on purpose thing is just strange. He acts like a fundraising email from ofa is a personal message from Obama. I saw a bleah debate performance and then Obama behaving like a normal candidate, going on the stump in swing states. How we get from that to Obama is throwing it is beyond me.
Chris
@burnspbesq:
“Battle for the soul of the GOP” my ass. It takes two sides to have a battle. The GOP has had only one side since 1994 at the latest. All the “battles” in the party since then have been purges.
Nicole
Please tell me the title of this post is a reference to that song from what, 1990, was it? I can’t remember who did it (even though, and I’m embarrassed to admit it, I had the song as a cassette single).
Matt McIrvin
@Hill Dweller:
Is it?
This is what people told Pew’s pollster they were. It’s well-known that many, many American “independents” are just as partisan as anyone, they just think that only tools identify themselves as part of a political party. I have little trouble believing that in the massive Romney-triumph atmosphere of, say, Thursday afternoon, enough of the Republican “independents” temporarily decided to call themselves Republicans to put the national fraction of Republicans at R+1.
Ed Marshall
It’s not Pew trying to put their thumb on the scale. They use a likely voter scale in their model that doesn’t change from poll to poll. They ask a series of questions that try and figure out if you actually a likely voter or not, and after a candidate has some sort of positive exposure in the media a percentage of people will start answering the qualifying questions as people with a serious intent to vote. The thing is, they wind up disappearing. Those people aren’t actually going to show up and they disappear back after a short time and the polls will go back to normal.
scott
@Matt McIrvin: Dunno about nothing. Obama went from being up 6 (which was already a bit of reversion to the mean after being up 8 earlier in the week) to being up 3.7 today. As I said, on the first day the poll could actually measure a reaction to the debate Obama’s margin took a 1.5 point hit. I’m not as interested in the Rand Poll for absolute accuracy as I am in direction, shifts, etc. It tracked Obama’s improvement after the convention pretty well and seemed sensitive here to the debate, although not in eye-opening terms. The fact that the post-debate hit seemed to be declining in almost linear terms over a few days seemed interesting. Could be BS, but worth looking at.
James E. Powell
What are the chances that anything anybody says or does in the VP debate will have any impact?
Valdivia
The tied gallup comes not from new numbers but from them creating a likely voter screen. Apparently though PPP for Kos has Romney ahead too. It’s going to be a long week.
amk
Can’t believe pew pee’ed on itself in such a manner. wtf happened?
Shawn in ShowMe
@James E. Powell:
The way I understand it, if Ryan “wins” then Mitt will go up by double digits. If Biden “wins” the polls will show a dead heat.
eemom
@John Cole:
Oh Cole. SO cruel. What if I was harboring a secret crush on you and utterly devastated by that exemplarily eloquent smackdown?
nepat
@Or something like that.Suffern Ace:
Doesn’t seem to matter. At this point The Debate has become like Godzilla in Tokyo, with Democrats and Obama-supporting independents (and whatever Sullivan is now) running through the streets screaming in panic. Sullivan is as guilty as anyone of starting the entire collapse in the media in the first place with his creepily emotional live blog of the debate reading like Scenes from a Marriage. The garment-rending from Obama’s side was a permission slip for literally everyone to freak out. Thanks, Sully, etc.
Ed Marshall
@Matt McIrvin:
Yeah, you got it. These sort of movements (and Obama’s convention bounce was similar) are mostly artifacts of the polling. When you feel good about your candidate you are more likely to pick up the phone, more likely to tell them you are going to vote for *sure*, etc..
Hill Dweller
@Mark S.:
It’s registered voter screen meant to resemble the demographic of the county. The nation is somewhere between D+5 to D+7, meaning more registered Democrats. In a national likely voter poll that drops to approximately D+2 or +3.
Pew used a registered(not likely) voter screen of R+1. In other words, they eliminated millions of Democratic registered voters.
As I was saying in the other thread, Ari Fleischer called this poll improbable. But the media ran with it without reading the cross tabs. ABC led their national broadcast with the news.
Jim, Foolish Literalist
@James E. Powell: ordinarily I’d say nobody votes for the second name on the ticket (Palin being a rare exception) but this race seems especially volatile, and I think the expectations game could be skewed by a lot of people thinking Biden is a ‘gaffe-machine’, and I think Ryan is thin-skinned and potentially weak. And does the media seem more highly strung this year, or am I watching too much Tweety?
piratedan
@Hill Dweller: and don’t expect a correction… after all, we still gotta push the narrative y’know.
Steeplejack
@John Cole:
Except that it didn’t seem to work, at least in the test earlier tonight. Or maybe DougJ and Mistermix limited the participation and didn’t explain that well. A lot of people couldn’t get in to comment.
Mnemosyne
@John Cole:
Is there a way to essentially rent more server space on an as-needed basis on the days you know the blog is going to go nuts? I honestly have no idea if that option even exists.
To extend scott’s metaphor, you don’t need that missile-equipped Hummer every day, but maybe you can borrow one for the few days that you actually do require one instead of trying to carry a full-sized sofa on your bicycle.
amk
@John Cole:
aka failing the moment.
Take a chill pill. Also. Too.
dollared
JC, I find myself agreeing with eemom. The purpose of the (*(@#*y%#(&*&$*# blog is to be here when we need you.
You’re like a Hooters that refuses to staff up during a college football coaches convention.
Matt McIrvin
@Hill Dweller: I could be wrong here, but I think Rasmussen is the only major pollster who actually re-weights poll data according to some model of party ID. If so, then Pew wasn’t jiggering the data to assume R+1, that was just what the people they asked told them.
I guess my personal philosophy about this sort of thing is more along the lines of Nate Silver or Sam Wang: I don’t go looking for ways to toss out weird outliers, in part because there’s a powerful psychological incentive to toss out only the ones I don’t like. I’d rather wait for more data to see what’s going on.
Of course, most of the political media have the opposite tendency, to pay special attention to weird outliers because they’re startling and exciting and maybe confirm some narrative. That’s wrong too.
dollared
@Mnemosyne: Yes. Eleventy hundred different ways. Hosters do it all the time. You can actually “burst” to Amazon or Microsoft Azure whenever your load exceeds your local machine.
Cole doesn’t like to replace light bulbs until someone falls down the basement stairs in the dark.
AliceBlue
@Valdivia:
Kos says that the majority of the responses in the PPP poll are from Thursday and Friday, the two worst days for Obama.
BillinGlendaleCa
@Mnemosyne:
Oh, that was you? I should have waved.
BillinGlendaleCa
@Steeplejack: Yup, never saw either of the two comments I entered.
amk
@dollared:
Didn’t he do that already ?
ETA: while being naked?
Mark S.
@Hill Dweller:
I should know this stuff, but do you mean they throw out a bunch of D responses til they get R+1? I thought if you made your sample big enough, you didn’t need to do crap like that.
But then again, I do remember reading on the pages of this fine blog that Gallup and some other polls were really over representing white people, so maybe I did know that.
Rita R.
@nepat:
So fucking this.
Mitt Romney could have blubbered incoherently for large stretches, vomited on stage and accidentally flashed his magic underwear and not one person on Fox News would have come close to the on-air freakout displayed afterward by those blowhards Chris Matthews and Ed Schulz.
Hill Dweller
@Matt McIrvin: It’s no surprise seeing them lowering the Dems likely voter numbers, but lowering Dems registered voter numbers to the point of Republicans leading +1 is laughable. That’s essentially eliminating 6 or 7 million Democratic voters.
Add to that some of the other shenanigans they pulled, and it’s becoming embarrassing for Pew.
A moocher
@dollared: in the dark? I’m surprised that makes a difference to his accident-proneness.
Hill Dweller
@Mark S.: Pew also over-represented white voters in the poll. Only 15% of their sample is non-white. The projected participation level of non-white voters in this election is 28%.
Every possible tactic to skew a poll Republican(race, geography, under-representing younger voters, over-representing older voters, etc.), Pew did it.
As I was saying earlier, Willard only being up 4 in this poll should actually concern his campaign.
Wiesman
I have some stupid questions and maybe they’ve all already been answered, but I did skim the 65 preceding comments and didn’t see them addressed directly.
And also, I know I’m going to get accused of concern trolling or whatever, but fuck it, I’m not a troll, concerned or otherwise, and if you don’t believe that, just pretend to believe it and please answer my questions anyway.
So…
1) We all correctly made fun of the Unskewedpolls.com guys who just decided to reweight their polls according to 2010 turnout, right? Why is complaining about the Pew sample any different?
2) I’m under the impression that party ID for the mushy middle swings according to whoever they have decided to support at any given moment, depending on the last commercial they just saw. Is that not enough to explain the Pew sample?
3) Assuming the Pew crosstabs have oversampled white Southern men, are people (including Cole?) actually alleging a conspiracy by pollsters to perpetuate the horserace?
4) Or is it that it’s a randomly bad sample but that the media (which DOES want a horserace, no doubt) is choosing to focus on the bad samples?
For the record, on an intellectual level, I’m not worried about the election because I think Obama is very good at getting elected (and even better at governing) and that he will win this election. On an emotional level, I’m going fucking crazy. Not Andrew Sullivan crazy, but kind of crazy.
I don’t believe in conspiracy theories and DO believe in Occam’s razor. My general reaction to the Pew poll is that it’s a combination of a genuine Romney bump, and a slightly bad sample, nothing more.
I know I’ll get flamed but I’d appreciate any actual answers.
Matt McIrvin
@AliceBlue: If so, it’s more evidence for my current wild-ass guess, that Romney had an enormous and genuine spike that put him well in the national lead… for about 48 hours. Following which Obama certainly didn’t recover his whole loss but may have gone back to a narrower lead.
And much of the craziness and inconsistency we’re seeing now is various kinds of temporal artifacting resulting from the inability of most polling mechanisms to deal with something that short-lived.
The thing that worries me is, the stories themselves have an effect back on the electorate through the media. The saving grace may be that most people just don’t pay as much attention to politics as we politics junkies do, so the meta-stories might not have as great an effect as people who read Andrew Sullivan would think.
Jim, Foolish Literalist
3) Assuming the Pew crosstabs have oversampled white Southern men, are people (including Cole?) actually alleging a conspiracy by pollsters to perpetuate the horserace?
This is, to me, the most glaring flaw in the poll. They seem to have cut the weight of minority voters in half, and if they have a reason for doing so, i haven’t read it.
SectionH
VA pres race looks a bit scary on Sam Wong’s site, but he still had O at 313′, R at 225 as of earlier today. http://election.princeton.edu/electoral-college-map/
Best graph of all they have, though,, is of the “who won the debate” going back to 76. (I posted the link to friends tht night, but can’t extract it from the Fb swamp, can’t find a search on PEC.) Basic message from the past: debates make no damn difference. The myth about Ronzo was just.
This is a srsly high-stake election for me this time: unlikely to get employed again at my age in a job which has insurance coverage, years before Medicare eligible, a family biz putters along but… I’m doing what I can, but so not watching the fucking hourly poll results. (PEC courtesy of Mr S, the insane optimist.)
Hill Dweller
@Wiesman: Pew has always been a credible pollster, which is why I think they deserve the benefit of the doubt. But the timing and horrendous sampling look suspicious, even if it was an honest mistake.
There was an 18 point swing with women(22 with white women) and a 19 point swing with voters under 50 for Willard. Good debate or not, that just seems implausible.
piratedan
@Wiesman: well based on the folks that have dissected the Pew poll, yeah, you could scream malfeasance but there’s a problem or two with that..
1) The R’s have done a great job about screaming conspiracy about everything from the New Black Panther Party to the BLS Jobs reports… my guess is, even if they (Pew) have jimmied the numbers, who’s gonna care when Republicans have already devalued the word with claim after bogus claim?
2) The media wants a horserace, it sells web hits and generates ratings, you think they’re going to pause and do a bit of self examination…. ummm no, it’s already front page on msn.com.
Matt McIrvin
@Wiesman: I think you’ve basically got it.
Also, something to keep in mind is that the percent MOE for the crosstabs in a poll like this get really, really large, because however small the whole sample is, the fraction of it that is, say, white Southern males is even smaller, and random statistical noise will affect it proportionally more.
If you slice and dice up an individual poll you can always go crazy either suspecting skulduggery or getting worried that something completely bizarre is happening. One of those randomly-messed-up crosstabs in a CNN poll was the entire basis for the claim that Prop. 8 passed in California because of homophobic black people turning out for Obama. Surprise, surprise: Andrew Sullivan made a big deal of it.
Wiesman
@Jim, Foolish Literalist:
@Hill Dweller:
I agree with both of you that the sample seems poor. I also think Pew deserves the benefit of the doubt — for their intentions, if not the results of this sample.
Taking a wait-and-see approach I guess.
Matt McIrvin
See also The Monkey Cage on paying too much attention to individual polls, either to tout them or to insist there’s something methodologically wrong with them.
kd bart
As far as the Pew poll goes, I would like to see a geographic breakdown of the sample from the previous one and compare it to this one. According to the sample breakdown, there was a large sample size from the South which is Obama’s weakest region and Romney’s strongest. Did the previous Pew poll also contain a large sample from the South or was the geographic breakdown much different?
FlipYrWhig
Andrew Sullivan’s whole embrace of The Bell Curve is based on the fact that he doesn’t understand what happens when people take data and draw conclusions from it. He sees the conclusions, sees some of the data, trusts that the conclusions suit the data (because that’s the part he has no idea how to assess), and accordingly gets played for a patsy over and over and over again.
Jim, Foolish Literalist
@FlipYrWhig: to put Sullivan in perspective, he invented his own brand of birtherism with his downright creepy obsession with Trigg Palin
eemom
On reflection, I do have to concede this particular point.
I mean it’s not like you’d ever use it, for example, for some joke of a “coverage” of a major political event that consisted of a series of dumbass “omg, ain’t she GORGEOUS?” photo ops.
James E. Powell
@Jim, Foolish Literalist:
I think any Tweety is too much Tweety. I know that he is allegedly/supposedly on “our side,” but I still can’t stand him.
West of the Rockies (formerly Frank W.)
Will someone please assure me that we BJer’s are not freaking out in the manner of Republicans who did not believe last Friday’s unemployment figures?
One theory we’ve read here today is that the media is whipping everyone into a frenzy to sell more copy. Another theory is that the polling itself was flawed. Another thought is that those mouth-breathing no-brain “undecided’s” have — at least temporarily — decided for Robmey.
I need a drink….
RadioOne
If Pew is right that there are now more GOP voters than Dem voters likely to vote, than we are toast. I don’t believe that, but obviously pollsters think the first debate depressed Democratic turnout signifigantly.
Joe Buck
Political identification isn’t stable; there are a bunch of people who will tell pollsters that they are independents one week and Republicans the next. Having the ratio of Democrats and Republicans in polls take wild bounces isn’t a sign that the poll is bad.
I’m not terribly worried as I expect the Obama team to wake up and do much better going forward (and to call Romney on his crap right away instead of just assuming that his reasonable, bipartisan, noncombative approach will impress anyone).
mdblanche
@freelancer: @Matt McIrvin: @FlipYrWhig: I’m noticing a pattern to some of
Denethor’s nonsense.
Brachiator
People are confusing polling with fortune telling. Childishly impatient, they want this stuff to be some absolute guarantee of what will happen on November 6.
Instead of trying to second guess the future, go see “Looper.”
It doesn’t make sense to talk about depressing Democratic turnout since voters haven’t yet, you know, turned out. And unless you have evidence of Democratic voters tossing their absentee ballots or stepping out of line at early voting sites, this kind of talk doesn’t even rise to the level of speculation.
But even if GOP voters and likely voters are more enthused, it’s still up to the Democrats to rally their side and get out the vote.
Polling is fun and interesting, but it’s not the final word.
Felanius Kootea
@Brachiator: Thank goodness for those early Dem votes already banked in Iowa ;).
fuckwit
Debates are a scourge upon what’s still left of our democracy.
They don’t accomplish shit. That’s been obvious ever since Reagan, hell ever since the press decided to crucify Al Gore for rolling his eyes.
They’ve certainly gone down the toilet since they got abandoned by the League of Women Voters, who clearly didn’t want to be associated with the fail anymore.
Shit, televised debates may have been useless every since Nixon and Kennedy. I don’t give a shit if the right guy “won” that one, they’re still toxic.
Televised debates are the worst forum for trying to decide who to vote for. And the people who make their decisions based on debates are the worst kinds of people to vote.
No real issues of substance get decided in a televised debate. There’s not enough time for that, and the format doesn’t allow for it.
It’s all high school popularity contest, pre-rational caveman which-alpha-male-has-the-bigger-cock kind of non-verbal bullsht. The voters who decide based on that irrational tribal primitive crap are destroying democracy.
Put the two alpha males in the ring, see which one stomps the other one! It’s gladiator contest– nothing scientific about it.
And then there’s the fucking infantile media. Earth tones, “there you go again”, rolling eyes, the fucking radio antenna wired into the shirt, all that trivial idiocy. The stuff they harp on is not just stupid, it’s dangerous.
This is serious. We’re trying to run the most complex society the world has ever known, solving difficult problems, using the world’s most powerful and dangerous technology. And instead of trying to actually understand stuff, we have people choosing their representatives and making critical, vital decisions that will affect the future– and whether we even have one– of the human race, based on who seemed “more confident” and “in command”.
Un be fucking lievable.
Cantrip
@fuckwit: This will be lost down this far, but you are right. Who are our pundits? Sarah Palin wannabes, mostly, who have a little cred rubbed off by where they happened to get into college, who constantly dream of being asked to the dance floor by the virile alpha male, who are constantly wrong but want you to swoon when they flaunt their starbursts.
But be of good cheer. If you really remember the 2008 narrative, Obama always seemed behind until you realized that he was counting delegates, building up his solid base (like a good community organizer should), and Romney would have to literally be the second coming of the Indian-visiting Christ to get the Electoral College to put him into power.
Patricia Kayden
@nepat: Love your comment. Made me laugh out loud with your Godzilla imagery. I usually love Sullivan, Matthews and Schultz, but have to take them to task for their hair pulling antics after the debate. Yes, President Obama should have done better, but it’s not the end of the world just yet.
JoyfulA
@muddy: I was busy trying to get an absentee ballot application last week, just in case my orthopedic surgeon makes me stay in this full-leg brace through election day. (PA is very restrictive on absentee ballots; I’d have the doc sign the form at my next appointment.)
Lo and behold, the Republican Party of Pennsylvania sent me two application forms in the mail–not to me but to the previous owner of this house 6 years ago. Thanks, GOP!
El Cid
If you were turning in research based upon surveys, say, to a journal or to your college professor or research team, their first task is to look at the question ‘Does this poll (as it was conducted) allow the researcher to answer the question it posed?’
You would do this of any research.
You don’t get to sit there and accuse the committee or reviewers as avoiding answers they don’t like by whining that the polls were skewed, unless the facts of the matter bear that accusation out.
So now I read that anyone who looks intently at the Pew results in order to ask themselves ‘What does this particular piece of research tell me?’ are now just whining that the polls are ‘skewed’.
Apparently, for reasons I don’t understand, major polls are now not just research tools to be examined as any other, but political institutions to be accepted.
rikyrah
someone else pointed out to me that they have been screaming all the time that no candidate can win the White House with less than the 31% of the Latino vote, and Willard’s in the low 20’s. how come that rule has changed
Judas Escargot, Acerbic Prophet of the Mighty Potato God
@rikyrah:
The rule changed, of course.
Just like that 8% unemployment threshold.
Matt McIrvin
@El Cid: The trouble is that if you suddenly discover this urge to pick the internals apart because you saw a result that displeases you, after having happily accepted all the numbers up to that point (as Democrats tend to like Pew), you’re probably going to deceive yourself through selection bias.
grandpa john
@Matt McIrvin: The screwiest part of the poll is that women went from +18 for Obama in Sept to even in Oct, that’s a change of 18 points in 1 month in a group that during the debate showed the most dislike for Romney
El Cid
@Matt McIrvin: Yes, but if you’re assuming psychological analysis, make sure you actually listen to the questions being asked.
The questions you ask of survey research aren’t “IS IT GREAT OR IS IT CRAP”, but “what can it tell us?” “What can it not?”
Let’s reverse the scenario:
Let’s say I did a poll. And then I publish it. And then I tell you what I think it says.
And you “pick apart the internals,” i.e., do your intellectual work and look at what is being claimed.
And then I say, ‘well, you’re just trying to deceive yourself’ no matter what your points are or how they are contextualized’.
For example, you could have a survey which comes to a conclusion, and a second survey which comes to a conclusion (or rather, which support a conclusion), and one can arguably be evaluated as supporting the conclusion whereas the other one does not.
Polls aren’t magical instruments which exist to merely inspire us, or frighten us — they’re supposed to be research.
They’re not magically floating above evaluation whereas work I would do at college or studies published in journals are subjected to the same.
It’s not “selection bias” to evaluate methodology.
And I have this “urge” every time I see a poll, because no matter the result — unless from time to time I’m just needling people on purpose — I want to know “what does this poll or survey tell me?”
Now, you could do meta-analysis, finding a statistical relationship between a certain poll’s results and some political outcome without respect to the “internals” (i.e., the actual data and results) of the poll, and that’s certainly legitimate.
But that’s not the same question as asking — again, as one would do of any poll — which questions does this poll address and how well does it answer any of those questions?
I’m not concerned with cheerleading anybody. But I’m also not into huffing the idiot pundit universe’s poll fumes any more than I do when they list what ‘experts conclude’ regarding something in foreign policy’.
grandpa john
@Wiesman: I’ll take your number 4 for the win. All polls sometimes get bad samples that do not truly represent the electorate. This pew poll is one of them.
grandpa john
@El Cid: Yes, while this poll may not be intentionally skewed and just be an outlier because of an unrepresentative sample, it is a flawed poll and does not represent the true demographic make up of the country
El Cid
@grandpa john: Even so, with enough information about what has been collected and how presented, it’s certainly possible to draw useful conclusions from one or other aspect of a given (assuming respectably conducted) poll.
If a poll accurately captures (as best as can be estimated) one or more demographics, then it can be useful to examine that. If it misses one or two, it’s also possible that this might or might not affect the overall result.
What happens is that in ordinary research you have to try to explain how those results affect each other. A given survey thus ends up neither as holy writ nor disreputable trash (per se), but one which might better address some questions and not as well address other questions.
Paula
Honestly? A guy lies his way through a debate and suddenly wins an election he was deservedly losing to a guy who’s clearly more competent than anyone his opposition has had to offer?
If that happens, the American people deserve everything they get. They had Bush Jr. as a model for electing a CEO warmonger president and the fact that a lot of them think it’s ok for it to happen again after only 4 years is a sure sign that the US is filled with a bunch of stupid fucks who deserve to get austeritied out of existence.
Romney’s performance was disgusting. But even more disgusting is the fact that he’s apparently getting away with it. So, whatever, if the ship is going down, it fully deserves to go down by the sheer stupid weighing it down.
scott
@El Cid: I agree with this. There’s a middle ground between uncritically accepting every poll on an equal basis, and extolling the ones you like while damning the ones you don’t. The earlier Pew poll I found slightly favorable to the President but not outside the realm of possibility, based on demographics. This one seems a bit out there not just on party ID (which is very fluid) but on others as well. Am I just a hack for looking inside the magic poll box to see what makes it tick? I don’t think so, and I add what I learn to the mix. What I got out of the Pew Poll is that there was a significant bounce for Romney, and I’m just skeptical it was that much. On to the next one.