Mark Blumenthal alerted his readers yesterday to what he calls the 'Gallup Crosstab Trove' -- the large amount of demographic data that Gallup collects and publishes on a weekly basis.
I took the average of Barack Obama's performance in the last three weeks of the Gallup survey and compared it against John Kerry's from 2004 as according to the national exit poll*. Obama is bettering Kerry's performance in all groups but one:
Obama's particular strengths are in the Midwest -- that's how a state like Indiana can be competitive this year -- among young voters, and among Hispanics**. The latter two groups are particularly interesting is they have far and away the most untapped potential in terms of improving turnout.
In the Democratic primaries, that potential was realized: the youth vote increased by 52 percent as a share of the Democratic electorate, and the Latino vote increased by 41 percent. But these are not groups that vote in heavy numbers traditionally, and so they may be among the first ones filtered out by likely voter models. They are also probably among the hardest voters to reach in surveys -- the youth vote because of the cellphone problem, and Latinos because of language barriers. All yet more reasons why polling is a dodgy and difficult business this year.
Oh, and what is the one group where Obama fails to outperform Kerry? Democrats -- although part of that may simply be that the undecideds aren't allocated in a pre-election poll whereas they have (necessarily) made up their minds by the time they take an exit poll. Kerry won Democrats 89-11; Obama presently leads by an average of 80-11 over the last three weeks of the Gallup poll.
-----
* In a couple of cases, the demographic groups don't match exactly -- Gallup uses slightly different age classifications than the exit polls did -- but we pair things up the best we can.
** Although, note that Kerry's exit poll figure among Latino voters is the subject of some dispute. Looking at state-by-state rather than national exit polls suggests he did a little better than this among Latinos.
TWO MORE NOTES: I am in meetings all day today, and so the daily polling thread is likely to be very tardy. And I deleted the other post that went up this morning, since you all clearly think I'm insane.
8.01.2008
Obama Outperforming Kerry Among Nearly All Demographics
@ 10:10 AM
Subscribe to:
Post Comments (Atom)
232 comments
Why SurveyUsa sleep since independence day? And why nobody poll Indiana?
This whole race is just waiting on that 10-12 percent of the democratic party that's just sitting on the sidelines. The question what's the holdup? Pumas? racists? just cautious voters? voters who just don't pay attention for the summer? I think overall it's good news that it's a handful of democrats who hold the election in their hands but I'd also like to know exactly why they're holding out.
I'm sure this is good news for McCain!
Damn. First I can't hear what Nate hears and then there is no post about something Nate can hear that I can't. Very Twilight Zonish.
;)
Wouldn't this comparison be more accurate if a Gallup poll from before the 2004 convention were used as comparision?
Interesting post, Nate. Obama's underperformance among Democrats compared to Kerry is obvious. But considering the number of potential factors contributing to it, it's remarkably small.
() The "Clinton voter" effect. Although Clinton supporters have been moving to Obama steadily, there are still wounds that have yet to heal.
() The "big tent" effect. There is a considerably larger Democratic coalition in 2008 than in 2004. The inclusion of former independents and a few Republicans makes it more difficult for any Democratic candidate to command the party unity Kerry had in 2004.
() The "Black" effect. It's been obvious for some time that the effect of racism on support for Obama is much greater among Democrats than Republicans. (The latter group has a number of alternate rationales for their opposition to Obama.)
() The "late decider" effect. The greater tendency of Democrats to remain uninvolved and undecided in Presidential elections (and to fail to vote in the end) has been widely ignored in light of the heavy participation in the Democratic primaries and the Obama "enthusiasm" effect. As those factors have been diminished in importance, especially in midsummer, the difficulty of reaching uninvolved Democrats has become more obvious.
The 10-12% of Democrats are most likely those who think Obama is too liberal.
On this date in history, August 1, 2004, in the Rasmussen Presidential Tracking Poll, John Kerry led President George Bush 51% to 46%.
For final results, please see Presidential Inuguaration, January 2005.
Can you PLEASE do an analysis of how the "likely voter" screen will be affecting interpretation of the numbers in states like NM, TX, and AZ?
It seems to me that if the screens' potential undercount of newly engaged voter is so significant enough that it may be responsible for, what? a 7 to 8 point difference in Gallup's national poll results, then it may be even more of an issue in these heavily Hispanic areas especially given how many of these voters are going to be undercounted by the "did you vote before?" screening exclusion.
Please.
Pretty please?
If Barack doesn't get his head in the game and out of the clouds, he's gonna lose. This is turning into a dog fight, and so far he is ill equipped.
Nate are these numbers? using NEW party ids? Because if that is so many of the democrats are former indpendents...
Otherwise:
LOOK WEST
LOOK WEST
LOOK WEST
The 10-12 percent of democrats are Hillary supporters. All polling suggests putting Hill on the ticket gives an immediate 5 point bump. If Obama can get those PUMA voters to vote for him he probably wins regardless of how bad a campaign he's running.
I think others above have hit the nail on the head with the "there are a lot more Democrats now" point. Between failing to count undecideds and the large conservative/independent population that has crossed aisles thanks to horrific Republican mismanagement, it's not surprising Obama's lagging Kerry in Dems, and I don't think it's worth much concern--especially given his strong performance with Republicans.
As a note, ooh, ooh, why is Nate insane? I missed it!
@Jeremy.
Right. Obama has a small lead at this moment only because he has only 75-80% democrats with him. Rasmussen has the same numbers.
Don't forget the factor primary.
We have the convention this month and Hillary will begin this month his campaign in the swing states for Obama.
McCain will win more democrats than Kerry(15%) but not enough.
At this moment the media speak about the fact than Obama is tied, but for me it's good for him with the fact than he has only 75-78% Democrats with him, and he leads.
Very interesting post, indeed. Actually pretty heartening given that the Democratic numbers should rise post-convention.
And I think jsh1120's point is probably right in that there's a portion of the democratic pool made up of former indies and republicans who dislike the GOP in general but who haven't totally come over to Obama yet. Thanks for this.
Ha. I don't think you're insane, Nate, I think you're eccentric. If you didn't know that already, there's no helping you. :)
If Obama does manage to outperform Kerry by eleven points in the Midwest, it is over for John McCain. There is (as the model is indicating) essentially no chance for John McCain to win the election without at least holding Obama to a draw in the Midwest.
Polling also suggests that when you add Nader into the mix Pumas chose him over McCain and it's McCain's numbers that fall not Obamas.
Nate, are you going to incorporate the regional crosstabs into your 538 regression?
The big question I've had for a while concerns all these hidden screens and filters and weights that various pollsters use. Systematic uncertainty in these probably overwhelms the statistical uncertainty.
Question on the youth vote/cellphone problem: A kid that schlepped him/herself to the primaries will almost certainly make the effort on Nov 2.
(Even if his/her candidate did not win the primary).
And you have the scale factor that corresponds to(+52%, but really you need to have it state by state).
So, do the weights applied to youth vote take this expected uptick into account? And what is the uncertainty introduced in the result due to whatever weighting they used? Should it be +52% +-2%?
or +-10%?
Same question about all the various demographics of course, but youth scale is so big.
Mr. Silver: When you do your regression fits, do the voter screens and filters used by the various polls (and the uncertainty introduced by them) affect the relative weight you give to any poll? Or do the "pollster rankings" account for this?
It's hard to believe just how much McCain is sucking wind right now. The Rass tracker this morning still shows Obama with a lead. How long has it been since McCain held a head-to-head lead versus Obama in one of the big trackers? Before the Obama nomination, surely?
Damn, I missed this morning's post. Someone please tell me what it is about.
My bet is that the Dem numbers shift once the convention starts, and Obama pulls closer to Kerry (maybe not surpassing him due to the primary race and racial concerns).
Why SurveyUsa sleep since independence day?
I hope they're busy preparing another 50 course meal.
On this date in history, August 1, 2004, in the Rasmussen Presidential Tracking Poll, John Kerry led President George Bush 51% to 46%.
That's an excellent point, a vivid example of post-convention bounce. (That was three days after the Dem convention had ended.) We will need to be careful interpreting the polling numbers in the weeks that immediately follow the two conventions.
First of all, this analysis is interesting, but it is slightly if not apples and oranges, oranges and clementines.
Second, so what was the big fuss about Obama not doing well with women?
Third, Bush won over Kerry by a small margin. If Obama is doing better in all these categories, it looks good to me.
Fourth, it is very early. Those Democrats, assuming it isn't an artifact of the system, will come around.
How about a map Nate? Can you have a graph of Obama vs Kerry
Blogs shouldn't have disappearing posts. Strikethrough with an update works best and maintains blog integrity. Whatever THAT means.
Tybalt,
Rasmussen numbers going back to June 4:
http://www.rasmussenreports.com/public_content/politics/election_20082/2008_presidential_election/general_election_match_up_history
Gallup data back to March 7:
http://www.gallup.com/poll/107674/Interactive-Graph-Follow-General-Election.aspx
Looks like McCain led Obama 46-45 in the Gallup poll conducted June 1-4.
I believe, btw, that Indiana bans robo-polling, hence it gets polled less frequently than other states.
Tybalt:
I just took a quick glimpse at Pollster.com's numbers and it looks like the last time McCain led in either Gallup or Rasumussen's tracking was in late May.
He led 47-42 in the Ras poll ending 5/28 and led 46-45 in the Gallup poll ending 5/27.
Guys i gotta be honest, I'm a huge Barack supporter, but I'd give anything to see Hillary tearing into McCain right now with that Clinton Attack machine all lubed up firing away. Obama fights like a girl. Or he doesn't fight at all. He needs to hire Wolfson.
Matt-
That is the reason Clinton was been the only Democratic President since 1980. Because he attacks like a Republican, and so does his wife.
Sure McCain hasn't had a lead in weeks, but he's behind by 1-2 points, and you guys are ragging on him about how lame he is. I bet the Gallup tracker has McCain ahead today.
Thanks Harvey. So if he's lost two a day since 5/28, that means McCain has lost 129 tracking polls in a row to Obama (Obama leads in today's Ras again, haven't seen Gallup yet). Excellent news for McCain!!!!
Something is wrong with these numbers.
It's not possible to improve by +10 among women and +6 among men, but +4 among married and +3 among single.
Everyone surveyed is either a woman or a man, so Obama's improvement should be at least +6. But they're either married or single, so Obama's improvement is at most +4. Inconsistent.
(Yes, there may be a handful who refuse to answer one of the demographic questions. But this is a huge discrepancy.)
--Scott
That is the reason Clinton was been the only Democratic President since 1980. Because he attacks like a Republican, and so does his wife.
VC makes an excellent point. Dukakis and Kerry fought weak campaigns (although to be fair to Dukakis, the concept of the all-out attack campaign wasn't thought to work until he got blitzed with it). Gore was a little better, but Bush gave him nothing to hit (and although Gore won the PV anyway, fat lot of good it did him, because he didn't keep the fight going).
Obama's gotta get down and rassle soon. I think he's been right to bide his time - McCain's punching himself out and it's only August 1st - but he's gotta do it.
I agree the nos look odd. Kerry got, what, 48% of the vote. With these ATB picks ups you would think Obama would be polling much higher than he is today. Something does not jive . . . .
Excuse me Matt, but many of us supported Barack in the primary BECAUSE he wouldn't fight like a Clinton.
I'd rather have a loser with integrity than a winner with none.
If McCain wins by taking the low road, it is only a reflection on the American electorate.
Nate, why did you delete the Britney post?
Matt J. H.:
I do have a problem with the way Obama seems to be attacking McCain's attacks rather than attacking McCain.
The template for attacking a multi-term senator is there; Clinton used it against Dole and Bush used it against Kerry. That's the kind of stuff that has worked and might work now.
If you get into a boxing ring and your opponent punches you in the gut, you don't stand there complaining about how he punched you in the gut. No, you punch the SOB in the jaw and knock him back a few paces.
Attacking your opponent may make you look like a bad guy, but attacking his attacks just makes you look like a whiner.
I'm with John. I'd love to know just how insane Nate really is, especially since I usually agree with him 100%. So really, I'll be finding out how insane I really am. Can anyone recap what this morning's post was about?
Well, you can blame Dukakis because Nixon had run all-out attack campaigns before. Either 1968 or 1972, depending on your view of things, was the first time a campaign like that was run.
The problem with Obama attacking is that he can be easily painted as "the angry black guy". I think that is why hes not responding too hard.
Never mind, I read the end now.
Here's a fun new poll out, for one county only (Maricopa County, AZ, the county that contains Phoenix).
McCain 43
Obama 38
According to the pollsters, McCain had led 53-38 in May, so he's losing support in AZ, although that support is not yet moving to Obama. Kind of parallels the national trends, where the most obvious feature of the last few months' polling has been that McCain has been dropping.
For reference, Bush won Maricopa county 57-42 in 2004, and 53-43 in 2000. In both cases, the margin was four points friendlier to the Republicans than the statewide numbers.
http://www.brcpolls.com/08/RMP%202008-III-01.pdf
My apologies in advance at singling you out Nick, but essentially Nick's viewpoint is why Democrats so often lose elections.
I remember telling Kerry supporters the exact same thing in 2004. For some on the left there's some glorified past period of politics where candidates took the high road, an age of election chivalry.
Hogwash.
Those who attack and attack well win. The Democratic Party used to know this, read your history. Read all about LBJ, Truman, FDR, Kennedy, and others, and how they did close to ANYTHING to win.
Your ideas, your policies, your goals mean NOTHING unless you are elected.
Granted I think Obama is smart enough to know this, or at least he has appeared to be. We'll see how long the gloves stay on and we'll see if he's another Kerry, or something else altogether.
It's not possible to improve by +10 among women and +6 among men, but +4 among married and +3 among single.
This is a fabulous mind-twister, but something tells me the numbers are right and the apparent discrepancy has to do with the fact that it's not Obama versus Kerry - it's the margin of Obama vs. McCain versus the margin of Kerry vs. Bush.
I'm sure some genius mathematician will have chimed in and explained it before I can hit "publish". Otherwise we need to say "okay... let's say there are 100 voters - 51 women and 49 men - and Bush got 30 men to Kerry's 19 ..."
"Angry black guy" is a risk he's going to have to take, because the alternative is "Typical weak Democrat."
For that matter, I'm not convinced that saying to John McCain "You've been in Congress for nearly three decades. What have you done to solve these problems?" gets Obama tagged as an "angry black man."
I tell you what Nick,
You can your honor and integrity, some of us need healthcare and jobs. Our kids are dying in Iraq, We're losing our homes and our jobs and the republicans don't give a rats ass about middle America. So you can take your honor and integrity and shove it up you ass, I need a president looking out for me.
Obama almost cannot attack, he chose and is forced to rely on defense. He runs with the "change"-brand. And that resonates with a lot of people. His attacks might be valid, yet the McCain camp will surely find a way to re-interpret these attacks as "Obama going the low road, he is just like other politicians". And there might be a critical mass of people who had enough of the old-style-politics.
Obama also runs as a unifier and he might have to rely on republican help to govern as effectively as he would like to (if the Democrats won´t win the 60 senate seats). The Clinton example shows how the divisiveness of the political climate can restrain the presidential course of action.
Harvey-
I'm not saying he wouldn't be smart to attack. I'm just saying he shouldn't put his name on it. Let some shady 527 do it. That way Obama avoids being the angry black guy, and gets to keep his high minded campaign.
If McCain wanted to touch the third rail of race, he should have gotten a 527 to do it (preferably one headed by a black Republican) and done it as a last salvo in late October.
ajbee,
thanks for the Maricopa poll.
I´m not from Arizona, but I wrote a six-page essay about the presidential race in AZ on a german blog.
If this poll is accurate, McCain is in danger. Maricopa will likely represent 60% or more of Arizona voters, and if the Latinos in Pima turn out huge, this race could be a Toss-Up- and this is important personally for me. Why? I´m working on a 538-like regression, and this regression showed AZ way too Obama- friendly all the time. OK, some of that may be, because the model does not recognize fully McCains home state advantage, but this Maricopa poll shows that my model may be better than the -whole state polls.
Ah! It's probably that a whole bunch of people have gotten divorced since 2004 (probably as a result of arguing about politics, based on the discrepancies between male and female!).
Obama is still down 21 points among white men, (slightly better than Kerry who was -25).
But I think that if it really begins to look like Obama is actually going to win, a fair number of those white men will start supporting Obama because a lot of them will want to vote for the winner. That phenomenon does exist.
Whoa, Matt, Nick -- cool it guys! :-)
I think there's a middle ground solution for you both. I agree that going negative is necessary, but the high road can me maintained if you stick to criticizing a candidate's record and staying away from character attacks or personal attacks.
Stick with policy -- but mention gut-instinct-simply-wrong policies that fit in a small soundbite. e.g. "John McCain talks a good talk on veteran's issues, but he voted against the G.I. Bill that would give them all a raise just to keep up with inflation."
Maintains integrity, mentions legitimate flaws, etc.
Without resorting to the "McCain has fathered a black baby" kind of crap.
SarahLawrenceScott--
It's possible. Kerry and Obama both trail among men. If the percentage of men surveyed increases, Obama can improve amongst both men *and* women without actually improving his overall numbers that much (and thus not improving his married/single numbers that much, as presumably approximately the same portion of men and women are married).
Example: Imagine you're in college, taking two classes, each with a Mid-term and a Final. In Class A, you get a 60 on the mid-term and an 80 on the final. In Class B, you get a 70 on the mid-term and a 90 on the Final. Your grade in Class B should be higher, right?
Not necessarily. If Class A's mid-term is worth 25% and the final is worth 75%, your grade is a 75. Meanwhile, if Class B's mid-term is worth 75% and the final is worth 25%, your grade is... a 75!
In this analogy, men are mid-terms, women are finals, and both married and single people are the grades. So I guess it's not perfect, but, y'know.
Guys, theres nothing wrong with attacking as long as its accurate, and on policy. McCain is going off the tracks because his ads are untrue and not about policy.
Obama can slam him every day on economic issues, do it with a smile, and tell the truth. He was doing it the first couple weeks after the Primary win, then he stopped.
I'm getting sick and tired of watching him run around like a little girl while McCain makes shit up. Sure Obama's been on vacation the last 6 weeks, where the hell has he been. He through away his 5 point lead. He has no clear economic message, and he's losing the media battle. Come on, its like amature hour. If this was any other year he'd be behind by 10 points.
This is turning into a fight, and he's losing the early rounds.
Rasmus, I'll be interested to see how Arizona turns out. I suspect it would definitely be in play this year were it not for McCain representing the state, but he also doesn't seem to be overwhelmingly popular there. He only won 47% in the primary...
VA Con, I kind of suspect that both campaigns want to put the race card back in the deck for now. There are ways in which race helps each of them, but they don't want to spend too much time talking about it (I've always thought that that's why Jindal won't be the VP, incidentally).
Note also that the issue came out at what's probably the low ebb of voter interest in the election, in the height of summer.
Married/single is not a binary. You might also be divorced, widowed, or living together/domestic partnership/civil union.
Please! I want to see the mystery deleted post too. It is getting far more attention as a mystery deleted post than it probably got as a goofy post
Wow when I came here and saw this headline I was shocked and still can't stop laughing.
Is this the best you can do to try to shine light on Obama's sinking numbers? Compare him to another loser?
538.com:
Electoral Projections Done Left
Nice Job Darian, rag on the guy. He puts this site together every day, It's the best anywhere in the world, I'm sure he doesn't make much money from it and you tear him down. Stop acting like a clown. If you want to attack someone, attack me.
I want to see the mystery post too.
Let's say there are 50 men, half of whom are married
Bush beat Kerry 26-24 overall Kerry wins 20-5 among the single (i.e. more intelligent) ones.
Fast-forward to 2007. Ten of the married Bush voters, being repressed Republican homosexuals, have gotten divorced. They still vote for McCain, but now they count in a different column.
Voilà ! Am I a statistician yet?
You actually made me laugh obsessed, well done.
It's funny how volatile intrade has been the last two days. Yesterday, Obama was down sharply and McCain up sharply, and now that's reversing, with Obama +2.5 and McCain -1.6.
It's actually pretty hard to figure out what moves intrade, but it's definitely not polling numbers.
Wow when I came here and saw this headline I was shocked and still can't stop laughing.
Is this the best you can do to try to shine light on Obama's sinking numbers? Compare him to another loser?
And yet Karl Rove has probably already read and analyzed his numbers.
Well, dang, now I really want to know what that morning post was about!
" Well, dang, now I really want to know what that morning post was about!"
Subliminal messages. Remember the 2000 election and the DemocRATS ad controversy? It was kind of like that.
VACon:
Of course, it's always preferable to let a surrogate (or a surrogate group) do the attacking. The candidate can then maintain a kind of bemused/disappointed "There you go again" air about him.
Maybe the Wes Clark incident was a kind of trial balloon for that kind of approach; it didn't really work, but Obama kept his hands relatively clean and it's still likely that such an approach would work on a more substantial line of attack.
Alex S.:
The entire idea of the "change brand" is that something's wrong with Washington; if McCain's "experience" can be used against him to portray him as typical of "what's wrong with Washington," I think it helps Obama far more than it hurts him.
In other words, if you can sand some of the luster off of McCain's "maverick" brand without doing much damage to Obama's "change" brand, I think it benefits Obama.
I think the surrogate argument was also why McCain began the campaign with a decentralized campaign structure, which otherwise seemed like a crazy idea. I'm guessing that the hope was that regional campaigns could attack Obama in ways they thought suitable to their local environments, while preserving plausible deniability for McCain himself.
I have a question. If Obama is doing so well with all these categories, how come he's up by 2 points? The McCain camp seems awfully agitated, and the Obama camp seems awfully confident. Is there some secret the campaigns know that we don't?
Is Obama further ahead and we don't know it? What gives.
@Matt
Look, I'm not asking Obama to lose, or to refrain from attack McCain's positions.
But to ask Obama to attack like a Clinton is asking him to bring up things like Cindy McCain's drug addition or Vicki Iseman, or John McCain's abuse of Cindy's jet or hell, just make something up.
Those things have no place in a political campaign.
My take - the best way for Obama to attack would be by one *particular* surrogate... a certain H. R. Clinton...
Matt-
The secret is the electoral college. The national number means little. Also, Obama is prety sure he is doing better than the polling numbers as he will increase turnout among the youth and blacks.
www.electoral-vote.com
I strongly agree with Virginia's Con advice of using a 527 or some other group. There is so much stuff that could and should be done.
Heck, if I were planning such a campaign I'd resurrect the Goldwater nuke commercial because folks have some unease over McCain's mental condition.
I agree with Stuart, is it time to break the seal on the negative campaign perfection? Hil and Bill are not just good at it, they are GGGRRREEEAATTT!!!
Hmm. That looks like pretty good news for Obama. It means, obviously, that he has a good chance of improving on his already strong polling numbers when the rest of the Democrats that he's not yet drawn in come home.
Although, to be fair, they may not come home at all. McCain still has some of that "straight talker" shine remaining. It may well be that those Democrats that haven't yet come home are the remaining Hillary deadenders who are bound and determined to vote for Hillary's polar opposite ideologically in order to punish Obama for the media's treatment of Hillary.
Odd logic, but it may yet keep McCain in this race.
nocool-
My wife is a Hillary supporter who will hold her nose and vote for Obama, but there are other older women I know who will not vote for Barrack as they have some weird need to prove sexism is more pervasive than racism.
We want the mystery post, we want the mystery post, rah rah. What the *#&@ did it say?
Scott, clearly the men/women is the correct number and it seems to go well with the other crosstabs. It HAS to be because Kerry exit was -1 (I think) and Obama Gallup is +3 so there should be a +8 margin improvement.
I don't see how the various theories offered so far about people getting divorced help one bit. The raw numbers are easy to find and seem at first glance to be about right. I think there's some bad luck with the rounding errors.
The nuke thing is such a cliche.
What the DNC should be doing is reaching out to those POW groups that hate McCain's guts and basically say hes a traitor. Its ridiculous, of course, but if they ran some slick ads like that it'd make McCain defend his war record and hurt him where hes strongest. Hillary would be doing this by now, no doubt in my mind.
McCain should get 527s to run ads about Obama's Chicago days (Wright, Ayers et al) as well as the old Democrat politicians there that don't like him.
Testing
The Cadence of Obama's Speech
Is it me or does Obama sound different when he is speaking to different audiences?
If you listen to the cadence in his voice when he spoke in MO the other day he sounded very country and southern, blacker, in fact, if you will.
Then when he is on national TV in front of the press corps or being interviewed by Fareed Zakaria, he sounds very professorial, Ivy League, if you will.
What's up with that? Mrs. Clinton did the same thing dropping her As and drawling like a southern belle when in the south or addressing rural voters.
McCain does not seem to do that.
Why the dissimilitude (dissembling?), I wonder?
Stuart:
Not just Hillary... prominent Hillary supporters too. There are some savage beasts within the old Clinton apparatus; turn 'em loose and let 'em have fun.
Matt J. H.:
The Dem advantage in party ID has got to be giving the McCainiacs some worries. It means McCain's got to make up a far larger gap than he would have in 2000 or 2004. In their estimation, going negative early is a calculated risk that could reap large dividends.
Yawp.
Let's say that you survey 100 people, 50 men, 50 women. 25 of the men are single, same for the women.
30% of the men and 70% of the women are for Kerry.
Your results are:
Men: 30%
Women: 70%
Single: 50%
Married: 50%
Now, Do another survey, 100 people, but this time 40 men and 60 women. Still half of each are single, Men are 30% Obama, Women are 70%.
Men: 30% (+0)
Women: 70% (+0)
Single: 54% (+4)
Married: 54% (+4)
So what we're seeing from the survey is that more married women and fewer married men were surveyed (or were declared likely voters).
PeteKent:
Senator Obama's only doing it to give you something to talk about. No worries.
Perhaps Darien would be happier if Nate added a couple of columns to the table which made explicit how McCain is currently performing compared to Bush in '04. Then the headline would be `McCain underperforming Bush ('04) in nearly all demographics'. I assume that would be good news for McCain.
In the Democratic primaries, that potential was realized: the youth vote increased by 52 percent as a share of the Democratic electorate, and the Latino vote increased by 41 percent. But these are not groups that vote in heavy numbers traditionally, and so they may be among the first ones filtered out by likely voter models.
This is perhaps the most effective indictment of the "likely voter" screen. We are left to assume that people who voted in primaries will not bother to turn out for the GE. Completely counter-intuitive. Democratic turnout in the primaries was staggering, and that increase was very largely due to Obama.
And dariencrow, you want to riff on "Obama's sinking numbers" though McCain has been behind him nationally for almost two months now. I've not seen any other electoral vote analysis putting McCain ahead of the "sinking" Obama, either.
As Colbert said, we all know that reality has a liberal bias.
I am completely unimpressed by Republican analysis based on 2000, and they seem to uniformly ignore the 2006 elections and the relatively low turnout for Republican primaries.
Calling all algebraically minded!!!
As we all know a tracking poll is a sore because it has n+2 variables (i.e. single day results) and only n equations (i.e. published tracking averages).
Today Rasmussen published his averages for days of the week since the end of the Primaries:
http://www.rasmussenreports.com/public_content/politics/election_20082/2008_presidential_election/little_day_of_week_bias_in_presidential_tracking_poll
(Without leaners. It is very interesting to find that Nationally, as well, Rasmussen believes his "without leaners" are the top numbers. He is unique in this belief).
This provides us with seven new equations so that the linear system is soluble.
The new numbers are even up to a decimal point.
Now, there's rounding error in the varios equations; we therefore can't take a simple linear equation solution and must use a linear regression (modeler, I keep learning!).
Once we crack this part of the code, we can use it extract the results from following days. (Soon of course rounding error would eat into our information, but that's still better than nothing).
Anyone willing to take this up? I honestly don't have the skills.
Virginia Conservative:
Obama has banned MoveOn.org and other 527s from attacking McCain as a person.
He also banned a California group from attacking Hillary a few months back. Obama believes in the spirit of finance reform and that's what he is trying to maintain.
But I do believe that his attacks are weak because they are at best subtly visceral.
Obama is walking to crossfire throwing knives just contrary to what he had proclaimed he would.
McCain is trying to shore up his demoralized republican base. That's why he will choose Romney.
MATT J. H. said...
I have a question. If Obama is doing so well with all these categories, how come he's up by 2 points? The McCain camp seems awfully agitated, and the Obama camp seems awfully confident. Is there some secret the campaigns know that we don't?
Is Obama further ahead and we don't know it? What gives.
Easy, he's doing well relative to Kerry, who lost. That doesn't mean he is winning by a lot.
I'm not the fondest of Hillary, but I think Barrack should pick her. Instant unity, 2 immediate attack dogs for Barack in Hill & Bill. The entire Clinton apparatus. I know she steps on the change mantle and Bill is a complication, yadda, yadda, yadda. Just win baby. After the last 8 years, I will forgo my better senses I just need a democrat in the white house for the country's sake.
Both candidates could go very negative through 527 and swiftboating.
Remember Swiftboat, publicly attacking Kerry in very vile manners.
Bush got on stage and said, how dare they attack Kerry in that way he was a decorated veteran.
Kerry's response, thats a negative attack from Bush.
Result, some believe the attack and Kerry's numbers go down, some believe the Bush response to the attack and Bush's numbers go up because he is so gracious to his opponent. Kerry has no proof Bush was involved in the Swiftboating, so his numbers go down, because his response criticizes the only person supporting him Bush.
Thats why this Vets for Freedom group may have a lot of power. Right now they are not specifically attacking Obama, although their commercial quotes him as being against the surge. If and when they step it up and outright say he is bad for the military, McCain can chime in and say how dare they, Obama has to respond in some way to the attack, if he blames it on McCain, we will have the same swiftboating scenario.
Someone asked for a map of the Kerry/Obama difference.
www.electoral-vote.com
has one. you just have to click back and forth between the 2004 map to see the comparison.
"It's not possible to improve by +10 among women and +6 among men, but +4 among married and +3 among single.
"
No way to know for sure but here's one scenario that actually fits those numbers.
Kerry Exit numbers: For every 100 voters, 54 are women, and 39 of those(72%) are married women. Of the 46 men, 24 (52% are married). Support among each group:
Single women: +23%
Single men: +15%
Married women: -4%
Married men: -34%
Now, in 2008, suppose you poll 100 people, and they're now split 50 male, 50 female. Now 30 of the 50 females (60%) are married, as are 24 of the 50 males (48%). Support among each group:
Single women: +30% [net +7]
Single men: +15% [net 0]
Married women: +1% [net +5]
Married men: -25% [net +9]
Not saying that's exactly what happened, that's one way this seemingly contradictory result could come about.
I jsut realized the Vets for Freedom site started today. They have an oped in the
Washington Times: Blind to reality, out today.
Matt JH-- seriously I think he's doing fine. If in fact his polls dip below McCain's for some period, then we can talk about contingency plans. This constant worrying that Obama is only ahead by 2 points is just silly. Look at the RCP chart from yesterday-- Obama's ahead in everything but tied in Montana and behind in Kentucky, Idaho, and Texas. There are like 10 other polls there state and national he's ahead in.
Besides "momentum" doesn't always mean anything with polls. They're all individual people making up their mind, and what we've seen so far is a very stable race at + a few % Obama, people really aren't making up their mind.
I'm certainly not counting any chickens, I just don't think it's necessarily the time to panic and make a "McCain is just one of the Golden Girls" ad.
Sorry to port this over from another thread, but I'd like the response to actually be visible. =)
James said...
Games,
Don't you think Democrats are generally more enthusiastic than Republicans? I meant no harm by this comment, simply that Democrats usually like to have their voices heard. Take the primaries as an example. In recent years, the Democratic primaries SLAUGHTER the Republicans in turnout, yet the general election is generally close. This to me suggests that there is a greater number of Republicans that skip out on the primaries yet vote in the general. [July 31, 2008 10:04 PM]
...
Seriously, do you just pull "facts" out of thin air and expect no one to notice? =)
Here's some data on primary/caucus voting in 2000 and 2004.
If you're saying that Democrats had a higher turnout in 2004 ... well, duh. The other party had a sitting president running for reelection, so the Republican primary was effectively uncontested.
Go back to 2000 -- the last time that both parties had a contested primary -- and what do you see? Oh, imagine that: Republican turnout is substantially higher, even in the early contests before Gore wrapped up the nomination.
So, compare that to this year -- contested nomination in both parties, but the Democrats vastly outperformed the Republicans even in the early contests before McCain wrapped up the nomination, with the obvious exception of Michigan and Florida. (Even with these two states, the turnout was almost as good for the Dems.)
The most extreme example I noticed: In 2000, the turnout in South Carolina was 9,657 D, 573,101 R. In 2008, the turnout was 532,468 D, 445,677 R. (In both years, South Carolina came before the nomination was settled for either party.)
Put another way:
* Republican turnout was down 22.2%.
* Democratic turnout was up 5413.8%.
Now, you're free to keep right on believing that Democrats are naturally more prone to participate in primaries and that everything we've seen this year is completely normal ... but at least try not to look surprised in November. =)
What do you mean, his attacks are weak. There are no attacks. Obama has not attacked McCain since June. He will lose if he keeps this crap up. He doesn't have to lie or be unfair. But he's nor even criticizing. one 30 second riff in a 30 minute speech is not an attack. Go after the guy. After the last 8 years of Bush, theres mountains of stuff to attack.
He'll be another great guy who just came up short on election day. And the democrats will wonder, how come we keep losing? Theres nothing wrong with the democratic party, but outside the Clinton's, they do not know how to win Federal elections.
Presidential elections are won in the media. it's about controlling the MSM beast. And right now, McCain is cleaning Obama's clock. The MSM feeds on outrage, and conflict and the outragous. McCain is feeding them well, Obama is starving them with the moral and high minded approach. The media tries to be fare and say things like "Why Axelrod did Obama go negative himself against McCain?" Of course he didn't , he's defending himself. But the media has to remain impartial, they can't tell the truth.
There is no truth, only 2 sides. And thus, it appears even though McCain has gone in the gutter, this week it will be Obama who played the race card and went too far against McCain. High minded crap does not work. The media does not care who's right or wrong, they just want a story. Give them a story Barack, give them something to talk about, or McCain will.
I'm not the fondest of Hillary, but I think Barrack should pick her. Instant unity, 2 immediate attack dogs for Barack in Hill & Bill. The entire Clinton apparatus. I know she steps on the change mantle and Bill is a complication, yadda, yadda, yadda. Just win baby. After the last 8 years, I will forgo my better senses I just need a democrat in the white house for the country's sake.
Apparently the states of Colorado, Oregon, Iowa, Wisconsin, and some others are too blue on Nate's map for your taste, 'cuz you're proposing taking a red paintbrush to them.
Wow - talk about shattering myths...
Where does Obama outperform Kerry the most?
Not with Blacks but amongst Latinos.
Not in the South but in the Midwest.
Not amongst postgrads but amongst "some college"
Not with Democrats but Republicans
On Obama not attacking McCain:
I'm pretty sure the Obama camp thinks that as long as McCain does't get above 45% or 46% in the polls (and he won't via negative campaigning), there is no need to go on the attack. McCain's partisan base is probably about 40%i.e. 40% that are impervious to negative attacks on McCain, so an attack style would only affect roughly 5% of his current support. Meanwhile, going on the attack would not increase Obama's numbers, probably just the opposite. For one thing, independents hate negative campaigning.
At this point Obama just needs to to keep developing the grassroots network and GOTV, stay out of the McCain gutter and keep making the general point that four years of McCain is four more years of Bush. There will plenty of time to attack McCain on Social Security, the deficit, taxes breaks for the rich, the flip-flops, abortion, and all the rest, should it become necessary. It may not be. Obama could come out of the convention with a consolidated party and a 10 point lead that he never relinquishes. It is much easier to govern if you have run a clean campaign that didn't include ubiquitous sniping and pot shots at the other side.
You people are impatient for separation from McCain because you salivate over the numbers all day long. But the numbers don't really matter right now. I think Obama is playing it just about right.
Matt J.H.,
I have to respectfully disagree. If Obama goes negative against McCain, I really think that it will cost him votes.
Negative attacks hurt the person being attacked, but they also hurt the person doing the attacking. Negative campaigning works if the former is greater than the latter.
The reason that Obama isn't winning by 8-10 points and well on his way to a landslide victory is because of undecided Democrats and Independents. In my opinion, the reason these people are undecided is because they haven't yet decided what they think of Barack Obama - is he an inexperienced far-left elitist w/ questionable associations, or is he a mainstream Democrat with some good ideas?
Attacking John McCain reinforces the former (as VA Conservative said, it makes him look like an "angry black male"). But attacking John McCain doesn't hurt McCain to nearly the same extent with these people because these people already have an opinion about McCain, and they're not going to be basing their vote in November on that.
Obama has to convince these people to vote FOR him, not against McCain. And you do that with a POSITIVE campaign, not a negative one. Certainly, Obama should point out the clear policy differences with McCain, but the key there is to not merely criticize McCain's position (and not McCain personally) but to contrast it with Obama's superior position.
It's August 1st and Obama has a clear lead by any measure out there. It's much too early to be overly critical of the Obama campaign. He's still the clear favorite.
Matt, I think we tend to forget that in many ways the ad war matters more than day-to-day campaign twists and turns, and that voter registration and get-out-the-vote activities matter more than either.
Given the whole combination of activities and events, Obama has been consistently and comfortably ahead. Yes, there's some risk that McCain's attacks will succeed in defining Obama, but two factors fight against that:
1). McCain wasted March-June in that respect, and that gave Obama four more months in which to solidify voters' impressions of him.
2). McCain still can't decide what line of attack to use. Two days ago, Obama was Britney, yesterday he was playing the race card, today he's raising taxes and hindering school reform (not an issue that plays to McCain's strengths).
In the primaries, Obama certainly showed the capacity to go negative when necessary, and actually had the rare trick of looking innocent while he did it. I'm sure his campaign is prepared to go negative whenever they think it appropriate or necessary.
Naomi, I hope you are right, but I see no evidence that you are. We have lost the last 2 elections because our guy did not know how to fight the battle properly. We are worried because we see the exact same thing again. This is the last 2 campaigns down to the tee.
Barack may be saving it for the final sprint after the conventions, but I do not see any capability thus far to battle the McCain attack machine. If this does not improve, he will lose.
GaMeS:
In 2000, South Carolina held a caucus on the Democratic side. That's why the numbers appear to skyrocket so tremendously.
----"On this date in history, August 1, 2004, in the Rasmussen Presidential Tracking Poll, John Kerry led President George Bush 51% to 46%.
For final results, please see Presidential Inuguaration, January 2005"
National tracking polls might actually mean something if we had a election on pure popular vote. Just in case you missed civics class, we have the ELECTORAL COLLEGE. State polls matter. So, to that device look up the state polls on Aug 1, 2004. You do realize that only one state changed from Kerry from Aug 1, 2004 to Bush on election day. Yes, the whole isea that Kerry had a big lead and lost it is pure fiction. Based on the elctoral college, the thing that determines the election he was never ahead.
Arnaud, SurveyUsa pools Washington post Independence Day.
McCain we´ll be out if he don´t put a conservative in his ticket. Because he´s a liberal republican.
The differential sampling solution to Scott's problem works in principle, but -
you really think there's an unintentional several point sampling gap between two polls each of which has N in the five digits?
more to the point, OTF, August 1 2004 was the height of Kerry's post-convention bounce. As Rasmussen themselves put it today:
Four years ago today, John Kerry led George W. Bush by four points (49% to 45%) and by five with leaners (51% to 46%). That matched his best day ever and was the last time he enjoyed such a sizable lead. At this point four years ago, the Democratic National Convention had already been held.
http://www.rasmussenreports.com/public_content/politics/election_20082/2008_presidential_election/daily_presidential_tracking_poll
So if Obama is up by five for even one day in the Rasmussen tracking between now and November, Jack's entire argument collapses.
Brad, electoral-vote isn´t a good page.
In my opiniom, RealClearPolitics is better and more realistic.
Yeah, don't worry, its in the bag. We'll take out 2 point lead in a year when we should be up 10. We'll beat them on the ground.
I'd take my 10 point lead and forgo the ground game thank you. I'm not suggesting that its over or anything. What I'm saying is that Mccain is running circles around Obama. Don't you watch the news.
When was the last time you saw some fake outrage on behalf of the Obama campaign. I see it daily from the McCain campaigh, like this "Race card" stuff. millions of people saw those clips and no one thought he was playing the race card. In fact. everyone believes he was telling the truth. But the media covers the "Race card" angle now for the second day. 2 days of anti-Obama media coverage in one ridiculous 200 word statement. Thats like $50,000,000 of advertising.
You see, I am seeing no evidence that Obama knows how to handle this. He doesn't accuse the McCain camp of anything. I HAVE SEEN THIS BEFORE. Don't worry guys, the voters see past this stuff. Sure they do. Where did Obama's 5 point lead go? Funny thing, Obama's lead disappeared the same day Schmitt took over the McCain campaign media message. Nothing changed in his campaign but their media message. McCain has been a terrible candidate and he's gaining ground. Why?
Because they are controlling the MSM. All this attacking and Britney Spears and the race card is the same as Hillary's kitchen sink. And you know what, sooner or later something sticks, then something else. Thats what we are seeing.
Its a slow erosion in Obama's positives and Obama does nothing. The only decent press Obama has gotten this month is his trip. If he had Hillary's media crew, He'd be ahead by 10+ points. It would be a low minded, scrap fest, but Obama would be winning big.
ajbeecroft:
If anything, the 8/1/04 numbers suggest that Obama's going to have a very good week in early September.
I remember that Kerry's post-convention bounce was a relatively small one; what happens if Obama's bounce is a little closer to average?
Gallup tracker is tied at 44-44
"So if Obama is up by five for even one day in the Rasmussen tracking between now and November, Jack's entire argument collapses."
Jack doesn't have an argument. He's just looking for attention.
Well, there isn't much evidence for Obama's positives going down, if you look at Rasmussen's numbers on favorability for the last two months. He's gone from 55 favorable, including 32 v. favorable, to... 55 favorable, inlcuding 32 v. favorable. During that time, his favorability has ranged from 53 to 59, while McCain's has ranged from 52 to 57.
In other words, not much has happened in the last two months.
http://rasmussenreports.com/public_content/politics/election_20082/2008_presidential_election/favorable_ratings_for_presidential_candidates
----"I have a question. If Obama is doing so well with all these categories, how come he's up by 2 points? The McCain camp seems awfully agitated, and the Obama camp seems awfully confident. Is there some secret the campaigns know that we don't?
Is Obama further ahead and we don't know it? What gives."
All these polls are using old models. They are greatly underestimating the young vote. 35% of people under 30 don't have a landline, it's 15% for those 30-45, for 45-65 it's 8% and 65+ 2%. So Obama's strongest demographics are being under polled b/c they can't be polled. Obama has already proven he can get them out in the primary and if you vote in a primary you are 95% likely to vote in the general. Obama crushes McCain in these groups that are being under polled. Obama has a few extra points in every state that isn't being polled. Also Obama is putting alot into voter registrationa and ground roots and get out to vote drives. LLet's alos remember the enthusuasm gap is wide. McCain is on life support and he knows us, thus is marked negative and deceitful ads the last 2 weeks. Obama 310 EC Votes minimum.
Obama is up by 1 in today's Rasmussen daily tracker. On August 1, 2004, Kerry was up by 5 points in the Rasmussen tracker and ended up losing by 2.5 points -- and he didn't have the Bradley effect to contend with.
In the primaries, the exit polls on election days repeatedly overstated Obama's performance so the networks delayed projecting easy Hillary wins that they thought would be tighter.
Thanks, all. I stand corrected. The numbers given here are possible. That's embarrassing, since my avocation is in a highly quantitative field.
It does make the numbers very hard for me to interpret, though. The two coarsest demographic cuts are women/men and married/singled, and yet a shift in the demographics of the sample or population surveyed amounts to about a 5% difference in the net shift. That means we shouldn't be surprised by shifts of at least that much in the other "net" numbers due purely to changes in the sample or population.
Hmm...actually, I bet a lot of this is the differences in when undecideds decide as a function of group. Let's try this scenario:
Half men, half women, half of each group married.
Election day, women go 70% for candidate A, men go 40% for candidate A. There is no marriage gap.
Three months out, suppose both married men and married women exhibit 20% undecideds. Single women are undecided 20% of the time, but single men are undecided 50% of the time.
That gives the following margins three months out:
Women: +32
Men: -18
Single: +9
Married: +8
And on election day:
Men: -20
Women: +40
Single: +10
Married: +10
Shift: Men -2
Women: +8
Single: +1
Married:+2
And yet there was no change in the proportion of each group voting for candidate A as oppoosed to B, or in the makeup of the sample. All that changed was that different numbers were taken out of the mix by being undecided.
My conclusion: although the numbers in the chart are valid, they're too difficult to interpret to be meaningful. What is an actual change in the opinion of a given demographic group? What is due to undecideds? What is due to a shift in the population (e.g. the aging of the population)? What is due to a shift in the demographics of the group sampled?
As I never tire of reminding people, on August 1, 2004 John Kerry had just finished his convention. I'm assuming you'll come back the day after Obama gives his convention speech and talk about how Obama is doing compared to Kerry in late August?
MATT J. H. said...
"I have a question. If Obama is doing so well with all these categories, how come he's up by 2 points? The McCain camp seems awfully agitated, and the Obama camp seems awfully confident. Is there some secret the campaigns know that we don't?
Is Obama further ahead and we don't know it? What gives."
I think in the areas that count (states, voter turnout, voter enthusiasm, donations) he is, Matt H.
A few things give.
1) Obama has led in every poll of registered voters for months. Mccain's ceiling seems to be 45%
Obama's floor seems to be 46% and whenever 3rd parties are thrown in, Obama swells to a 10 point lead
2) Just as Obama figured out prior to Iowa that it was all about delegates, and completely out-outmaneuvered Hillary as she invented 7 different rationales for why she should win, he is well aware from 200 and 2004 (don't forget that Kerry was 60,000 disputed votes from taking the election despite trailing by 3 million voters) that national polling is irrelevant, and he has many paths to 270
3) McCain and Obama are well aware that there is a 30% gap in enthusiasm for the two candidates. McCain's own base is lukewarm about him, and the more he demonizes Obama to feed his base and the closet racists, the more he turns off independents who remember the 2000 straight talk McCain. That McCain has been replaced by the Sour Grapes Express and/or the Low Road Express, take your pick. That is a turn off for indys.
The only unknown known is whether the corporate media's reflexive parroting of the McCain attack line of the day (Rachel Maddow, Olbermann and surprisingly Andrea Mitchell being noble exceptions). The media played this game in 2000 with Gore the liar and 2004 with Kerry the flipflopper above the surface, and swiftboating below the surface. Hard to say if enough Americans can be played to make it 50-50 on election day, so that some states can be stolen.
A few differences: Ohio machinery is now controlled by Dems, so harder for the GOP to steal that one, and Charlie Crist is no Jeb Bush, so it is possible they may more or less count those votes.
A Pew poll of substantial size came out today and echoes a CNN one that came out a few days ago, 47-42 (CNN 51-44), so I would take the national trackers with a grain of salt.
Lastly, great catch Tybalt. The post mentioning the Kerry 51-46 lead was 3 days after his convention, where Kerry should have been at his high water mark). The GOP strategy under all Bushes has been (and McCain camp is now run by Bush-Rovians) is to use August as Dem demonization month. Happened successfully with Dukakis, Gore and Kerry and Not so with Clinton 1992. The Dems were smart to move their convention back to the end of August, after the slime attacks, since the contrast with McNasty and the other old relics immediately following will be striking.
Interestingly with both Gore and Kerry, by the end of October, the Dems had made up the ground of the slime attacks, and no one should forget that Gore did win in 2000 by any measure other than that used by Antonin Scalia.
We will see what stuff Obama is made of. He is a singular candidate, and youth, Af-Ams and Latinos will be out in numbers that the likely voter models can't conceive of. I think he needs to continue his John McCain can do better than that, and honorable man running a dishonorable campaign themes, because the ugly attacks and responses of Tucker Bounds, Rick Davis, Steve Schmidt and McCranky himself feed into that image.
Bottom line, yes. I think the McCain camp knows if they cannot keep it close now, it is over, and as it is, they have to hit an inside straight to get to 270. The desperation is palpable, and I think they may end up turning off even the lapdog press. Jon Stewart is a pretty good barometer on this, and he did as rough a piece on his buddy McCain as I have ever seen him do yesterday, starting with "dick move of the week" and going from there.
Stay tuned
Obama confirms he is a socialist:
He announced that Oil companies will fund a $1000 check to low income families, if he is elected president.
http://www.politico.com/news/stories/0808/12237.html
Obama the Socialist at it again! Don't know where in the constitution it says that low income families are entitled to $1,000 handouts from the governemnt taken from oil companies.
Don't know where in the constitution it says he can't.
"Obama confirms he is a socialist:"
Tyrone,
Cutting into windfall profits with a surtax is not socialism.
If Obama nationalizes the oil companies he will be a socialist.
When that happens, get back to us.
I for one think its unconstitutional to tax a company just to give that money to low income families. The constitution justifies taxation by representation, and that is not taxation by representation AT ALL.
----"Obama confirms he is a socialist:
He announced that Oil companies will fund a $1000 check to low income families, if he is elected president.
http://www.politico.com/news/stories/0808/12237.html
Obama the Socialist at it again! Don't know where in the constitution it says that low income families are entitled to $1,000 handouts from the governemnt taken from oil companies."
Please go back to Jr. High and learn the defintion of socialism.
McCain and Bush are for corporate welfare. We subsidize the oil industry to a tune of 18 billion while they owe the US government over 9.5 billion in unpaid royalties. The Congress should say drilling or new leases isn't even an option till you pay your debts. Bush won't enforce the royalty payments and the people in the GAO that brought it up were demoted, told to leave it alone, and harrassed, forcing some to resign.
Btw, why do we have to subsisdize a industry with record profits? Corporate Welfare!!! You don't think that it was coincidence that McCain flipped on off-shore drilling and he received 1.2 million in oilk industry contributions that month. McCain is in the pocket of oil as much as Bush
Scott, the difference is so real it has to stand for something which actually happened in the world, so it's worth thinking about in more detail.
The one obvious difference is that 2004 is a self-selected group of those who decided to participate in the 2004 elections.
It may also be that Gallup deliberately oversamples certain groups so as to get finer-grained results with them.
What do you think happened here concretely?
Also, did you see my note inviting people to crack the Rasmussen code? It would be a pity if no one takes that up.
Good post, Michael!
Another point to keep in mind... for all the vaunted brilliance and invincibility of the Bush-Rove machine, it was never remarkably successful in presidential elections, having won in 2004 by 2.46%. And we all remember 2000. By historical standards, Bush's electoral college victories were quite small, too.
All of which means that McCain has to do very nearly as well as Bush did in 2004 in order to win at all, and in order to have any degree of safety he needs to be doing better.
So far, he's clearly doing much worse. In 2004, the lead swapped back and forth all the time in the polls; this time, about one poll in 50 has shown McCain ahead. That one USA Today/Gallup poll (McCain +4 LV) excepted, no poll has put McCain up by more than two points since early May, and only 5 polls have had him up by 5 points or more since Feb 1. Obama, by contrast, has been up by five in mulitiple polls in July, and has been up by as many as 15 points in some polls.
Of course either side can win this election, but McCain will need an extraordinarily good campaign, something he's pretty obviously lacked so far.
http://www.youtube.com/watch?v=ogTas21jZ6M
I love the socialist argument. Thats funny. Cut my $1000 check from Exxon Mobile, call me a socialist too. Those Oil companies are bastards.
Obama at what appears to be a historic low in the Gallup tracker. I think you have to go back to late march early April to see him at such lowe levels. meanwhile McCain is nearly at the top of his performance.
Are we in the midst of a trendline or an inflection point?
Where will be on Monday after news of the week is dissected all weekend?
Still tied?
Obama + 3?
McCain + 3?
Obama needs to pick a running mate and soon.
Mrs. Clinton is watching and waiting.
Thanks, ajbee
You are right, and of course, Bush's "win" in 2004 (we will never really know what the hell happened in 2004, since Kerry didn't have the testicular fortitude to fight it) was the lowest percentage and electoral college win by an incumbent in the history of the country, aided and abetted by a bad candidate and a complicit media carrying the story line.
Look at the RCP electoral college map, which you can tinker with to your heart's content. according to them, McCain is down 238 to 163 with 107 "toss-ups" (some, like NM, fairly dubious). It looks like McCain has to have a perfect storm: run a perfect campaign, hope for voter suppression, have the media spin his Obama the scary exotic vacuous elitist presumptuous uppity one story line non-stop, hope to God that all the youth, Af-Ams and Latinos DON'T vote (see voter suppression) and still hit an inside straight. Obama can afford to lose Florida, Ohio AND Michigan and still have many paths to 270, whereas McCain needs all three and still some fortunate breaks/chicanery.
As Blumenthal for pollster.com pointed out today:
"Alan Abramowitz notes that Obama has consistently led in national polls over the past two months. In fact, according to national poll results listed on Pollster.com, Obama had been tied or ahead in 50 consecutive national polls through Sunday. Sure, many polls may show Obama holding a lead within the statistical margin of error, but if Obama and McCain were actually tied, we'd expect as many polls showing McCain ahead as show Obama ahead. Based on some basic calculations, the probability that 50 consecutive national surveys would show Obama tied or ahead if the candidates were actually tied is .0000000000000009. In short, this race is not a "statistical tie," despite what a few scattered surveys (drawing disproportionate attention from the pundits) indicate."
Not true--McCain has never been over 46% in the Gallup tracker since the beginning of June. And the race has been tied a couple of times in July. True, Obama is down, but undecideds are up. I think these are some Hillary supporters who are disappointed that she won't be the VP nominee.
Lol, good try Pete
Humanist: Since there are 5 more equations than unknowns, something like singular value decomposition might be applied to determine the results from individual days (from the three-day average).
This, incidentally, is not the first time Obama and McCain have been tied at 44 in the Gallup tracker; they were tied at 45 on 6/25, then tied at 44 on 6/26-27.
"McCain is nearly at the top of his performance"
Interesting observation "Pete Kent."
Are you saying McCain is only at the top of his game when slinging mud?
I continue to hope McCain will pick Mitt "flipflop" Romney so NC will go blue.
Gregm, that's why I was flailing my hands about me so violently and screaming at the top of my lungs - because I noticed we have more equations than variables!
Could you please do a singular value decomposition for us? I wouldn't know how to begin. I never did any mathematics beyond set theory and logic, and statistics I did not do at all (I know its pre-20th century history reasonably well, but that's it...)
It's way too early for people in the race to be panicking about anything or to be crowing about anything. But the facts are pretty simple and, as an Obama supporter, are a little worrying:
(1) Republicans know how to fight and do fight. Obama needs to take off the gloves now or he will turn into another Dukakis.
(2) Although I find the idea of a Democratic Indiana (my home state), Montana, Missou, and Fla.,to be heartening there are some serious doubts in my heart of hearts that this will happen on voting day. Indiana barely voted for Johnson when Goldwater was running and only voted for Roosevelt once.
(3) If an experienced person (Hillary) or (I hate to say it but I'm going to) a person of Caucasian extraction were running this race would be locked down right now. The longer that Obama appeals to the mind and not the heart or the wallet, the more trouble he is in.
(4) I'm personally waiting for the Weekly Reader poll. I was all charged up about Kerry and Gore until those came back for Bush. That's only half a joke.
(5) To be redundant, I'm from a semi-redneck state and live in a redneck area of California (the Central Valley). I know how much racism lurks in the hearts of Americans. I am very concerned about the political correctness of polling results -- that is, I do believe that people tell pollsters what they think is acceptable and will switch when they get into the voting booth. Call me a 40-year-old cynic, but there it is.
(6) African-Americans, Hispanics, and Young People don't vote. Obama can't count on them and shouldn't. It is a mistake to assume that African-Americans will turn out in droves because Obama is black. Everyone is touting the Hispanic vote but I still remain a skeptic of its effect, even though CNN gave them room at the table during the Obama/Hillary debates.
(7) Obama has to choose Hillary, which may lead to his ultimate defeat. Or ultimate victory. It's a paradoxical position to be in but he needs that boost, nobody else will give it to him.
(8) In short, in my expert-idiot opinion, Obama is where he needs to be but not where he should be. Neither campaign is running a great campaign and one of them is poised for a knock-out punch. All it takes is one bad f'up and that's that.
As far as commentors go, this is the one site where I have seen Obama supporters (like me) who actually get it.
Obama is going to lose this thing unless he gets into the alley and cut the old hunchback boil face up. All the other sites have people sitting around saying "wait until this or that date" and things will get better but that’s not the case.
Voters don’t care about substance at this point, they want to see who has the tenacity to get in there and fight like hell. They want someone who will do that for them as President. Right now they see an old man who will do ANYTING to win, and another who just sits there when he accused of wanting his country to lose a war for his own political benefit.
I agree that McCain is a scumbag for doing it, but so what. He's tied in Gallup now, and its getting worse for my guy by the day…..
Why are you laughing Matt JH?
Michael: Your electoral analysis is sound. I think McCain will have difficulty in winning a close election unless he finds some sort of regional wedge issue that will re-secure the West for him and nail down the heartland. I don’t really see him doing that.
I think his strategy is more proportional -- a rising tide lifts all boats -- and that he is playing for an across the board victory.
I have said before he will either lose in a close election with Obama or win with an electoral college landslide. Maybe even as big as the one against Dukakis, but certainly with numbers approaching Clinton's.
That will require that he drive up Obama's negatives, dispirit the Kids by showing them that he is just last months flava and giving the rest of the country sufficient reason to pick him over the supposed change agent.
By election day we will be a dramatically different country than we were in January.
The War will be seen as won or winnable and the stakes there sufficiently great and the risk low enough that the people will want to consolidate that victory.
Gas prices may yet continue to fall and even if they don't Obama's opposition to drilling and his party’s queer allegiance to being Green will cause him fits all Fall. Finally, the economy is showing signs of turning over for the good. GDP was up again for the second consecutive quarter -- things are moving in the right direction. While unemployment rose today, employment is a LAGGING indicator of recovery and we seem to have enough time to get in a few more months of positive economic statistics as to make the case that prosperity is just around the corner.
At that point since things will not appear to have gone or be going to hell, will the people really want to roll the dice on Barack Obama? Will they need to?
I doubt it.
McCain will win. He will take MI, OH, FL and might even get NV back. He will hold the south and the rest of the traditional Republican areas and may indeed surprise elsewhere as he and his resurgent party get a thoughtful second look.
Since people are obsessed with national polls lets make this clear for all those who skipped Jr. High civics class or were suspended that day.
WE HAVE AN ELECTORAL COLLEGE SYSTEM!! NATIONAL POLLS MEAN NOTHING! STATE POLLS MATTER!!
The density of people is like lead!
Humanist: I should have also mentioned that SVD is just one way of doing a linear least squares analysis, so other ways should work in principle. Is there somewhere all the Rasmussen data is compiled? I might not get a chance to work on it for a few days. One concern is that since the days-of-week data is so close together in value, it might not be providing all that much new information, so the problem may be ill-conditioned.
Pete said,
"McCain is nearly at the top of his performance""
Polling at 44% means he is at the top of his performance? Yikes. This election is over.
When McCain can get to even 48% on a consistent basis you may have something to crow about. Until then, you've basically got nothing. All McCain has done is drive up Obama's negatives. His own support is where it's always been, and always will be, because there is an unmistakeable majority in this country that detests Bush so much they will never vote for a Republican to succeed him.
Learn to accept your 44-47% Pete, because that's all you're gonna get this time.
I agree "Operation Infiltrate." Its time for Barack to bring out the South Side of Chicago guns and tear Grumpy McBush apart. It can be about issues, but it should be an unrelenting avalanche of economic attacks that makes McCain look like an economic Hitler. I believe Obama knows whats going on, which is good. We need to see how he reacts.
State polls only matter on election day. National polls always pick up trends before you can poll in enough states.
It has been shown that you can't lose the popular vote by more than about 1/2 of 1% and still win the electoral vote count.
Everybody, please check out Drudge link to a ad about "the chosen one"
I disagree that national polls do not matter at all. If anything I think the Idaho poll last week has shown us how national polls and state polls can work together to get an idea what is happening in states that are not frequently polled.
To Pete Kent:
To me it always seems like you go between realist and dreamer. I agree with some of the problems the Obama camp might have but talking about a "resurgent" republican party is an illusion. I have seen absolutely no evidence for this and I have no idea where you come up with this. They are about as resurgent as the San Francisco Giants: Hanging in their better than anybody expected but terribly uninspiring.
Obama's strength and weakness is his popularity.
The media talks about Obama constantly, because that's what people want.
Obama tried to put the focus on Bush/McCain by saying nobody is asking him anything controversial but he needs to do this more consistently.
Other than McCaskill and Wrexler Obama's TV surrogates are LAZY.
The data are all here:
http://www.rasmussenreports.com/public_content/politics/election_20082/2008_presidential_election/general_election_match_up_history
http://www.rasmussenreports.com/public_content/politics/election_20082/2008_presidential_election/little_day_of_week_bias_in_presidential_tracking_poll
I am pretty certain the days of the week data are without leaners (this should be easy to verify).
I see what you say about the low information in the days of the week. Interesting to see how statistical thinking is rather different from linear equations thinking, even though on the surface the problem appears so linear-equations-like.
It would take a little bit more work, but one could also do the same with gallup, taking their crosstabs and extracting from them the raw weekly averages, and then combine those with the 3-daily averages.
The simplest of course would be for the tracking polls to PUBLISH their daily results. Why don't they?
I also wanted to tell OTF, while I'm here, that respondents to National Polls live in States.
MATT J. H. said...
Guys, theres nothing wrong with attacking as long as its accurate, and on policy. McCain is going off the tracks because his ads are untrue and not about policy.
Obama can slam him every day on economic issues, do it with a smile, and tell the truth. He was doing it the first couple weeks after the Primary win, then he stopped.
I'm in total agreement here. If you let your opponent attack you with untruths and don't answer back, then you can't blame the public for believing untrue things about you. And it's not underhanded to point out those untruths and go after your opponent's credibility. If your opponent lies, there should be a price to be paid for it. Then maybe we can have a discussion that is more fact-based--precisely what Obama is calling for.
And McCain has been wrong on so many things: why Obama didn't visit hospitalized troops, Obama's responsibility for gas prices, Obama's willingness to raise "your" taxes (unless "your" means only those making over $250,000), claims that slightly higher taxes will kill the economy (despite the clear counterexample of the early 1990s), the effectiveness of a gas tax holiday as the solution to our oil problem. And most recently an ad questioning Obama's leadership that is totally based on images and doesn't contain any content whatsoever to back it up.
There's nothing undignified with pointing out that your "straight talk" opponent is not talking straight. Then the conversation can turn to things that really matter.
Matt j.h.,
I agree that Obama needs to attack, but he needs to attack on an emotional level.
Obama's attack 527 groups should run ads showing McCain's numerous flipflops on issues and ask:
Can McCain be trusted?
This could work as the election is also past the "John McCain war hero" phase.
Obama has mentioned how McCain got big bucks from Big Oil to flip flop on drilling. Another good ad would be
Who does John McCain represent? You or Big Oil?
Obama has to find a counter to McCain's disingenuous lead "narrative".
It has been shown that you can't lose the popular vote by more than about 1/2 of 1% and still win the electoral vote count.
Nonsense. It's very possible that Obama could drive up the popular vote in all of the southern states, but not quite enough to win any of them. If so, McCain could win the electoral vote count even while losing the popular vote by millions.
NC Moderate:
Absolutely correct -- emotional level = stop talking to the intellectually engaging stuff and start talking to the heart.
Regarding the whole national polls debate, I find this to be an argument used by people who have opinions that are swayed by the pundits, who don't really have much substance. Yes, we obviously have an electoral college system, but people are in the sway of the 2000 election. There have been exactly two elections in which the popular-vote loser has lost. National numbers drive states, just like a previous author wrote -- they all live in some state somewhere.
I like Naomi's comments usually but I have to disagree with her -- this election is not over.
And as smug as you'd like to be, PeteKent, you should be worried about your candidate. There's a little too much cause-and-effect thinking in your thought process this early in the campaign. Any time that you talk about maybe rescuing Nevada for the GOP, you should be worried.
humanist (again):
"The simplest of course would be for the tracking polls to PUBLISH their daily results. Why don't they?"
The MSM and parts of the blogosphere already go crazy trying to explain statistical noise, PIE, and the like as the effect of some news event. Unsmoothed tracking data would make that even worse.
This post is a revised version of my earlier post with a number fixed.
Humanist: I tend to agree that it's something that happened in the real world, but the number of possibilities are still too large for me to guess. They fall broadly into two categories:
--the demographics of likely voters has changed, partially because of turnout/registration changes, and partly because of changes in the population as a whole
--different demographic groups decide at different rates
Trying to untangle those effects from these numbers is a gordian knot.
What we need is Gallup numbers, broken down by demographics, from an equivalent point in 2004. Of course, there is no "equivalent point," since the convention was held earlier. But at least it would do better with the undecideds.
By the way, one effect of undecideds on these numbers is easy to see and understand: typically, spreads on election day for a group which swings mostly one way will be larger than spreads in an early poll. For example, Democrats might by 70% for the Democratic candidate, 10% for the Republican, and 20% undecided early on...but that means that 88% of those who have decided are going for the Democratic candidate. If the same is true for the undecideds (dubious, but a good starting point), then the lead among Democrats "goes up" from +60 to +76. In my mind, that is not indicative of real movement toward the candidate.
So let's look at the numbers in this post and see what groups are like that:
Men were -11 for Kerry. Since that group went against Kerry, we should expect undecides to make it look closer now. Obama is at -5, so that checks.
White men were -25 for Kerry; -21 for Obama now. That checks.
White women were -11 now; -2 for Obama now. That checks, although it's almost surely not just the effect of undecideds.
The East was +13 for Kerry, but is +17 for Obama now. That's against the effect I'm describing, so probably represents a real difference.
The South was -16 for Kerry, and is -8 for Obama now, in concord with the expected effect of undecideds.
Whites were -17 for Kerry; -11 for Obama now. Another check.
Age 18-29 is a true plus for Obama, and a big one; the effect of undecideds should have made the margin bigger for Kerry, not smaller.
Blacks once again defy the expectation; Obama has a real improvement there.
Hispanics are the shocker: an overwhelming real improvement for Obama.
Postgraduates also show an improvement that goes against the expected effects of undecideds.
Democrats: +78 for Kerry, +69 for Obama; I bet that's largely the undecided effect.
Republicans likewise can at least partially be explained by undecideds.
As can married people.
But singles once again show a real improvement for Obama.
Summary: Obama is "really" doing better than Kerry in the East, with young people, with African-Americans, with Latinos, with people with post-graduate educations, and with single people. He's also doing better in a lot of the groups that split near 50/50 (middle aged people, the Midwest, people with less education), but it's hard to disentangle that from interactions between demographic groups.
Something that always puzzles me is the way people take very short-term trends and extrapolate them all the way to November. When Obama was six points ahead in June, it was a mistake to think he'd keep increasing his lead throughout the campaign, and it's a mistake now to think that McCain's got the upper hand just because the tracking polls have dipped slightly.
In reality, the tracking polls fluctuate every day. Just as you can't predict how the stock market will be doing in November based on whether it went up or down today, you can't predict the election outcome based on a day's tracking, or even a week's tracking.
For the last two months, the polling has consistently shown Obama with a 2-6 point lead. The best polls have shown Obama ahead by 10 or more, the worst have shown him tied. If you look either at Nate's Super Tracker, or at the pollster.com averages, you can see that the overall trend, to the extent that there's been one, has been neutral-to-positive for Obama.
Of course McCain could come up with a very clever attack strategy that brought Obama's numbers down drastically, and maybe he will yet do so. All we can say for right now is that he clearly hasn't done so yet.
Humanist: I just noticed a problem that prevents this from working:
Rasmussen says in the article about the averages of the different days of week, "Keep in mind that the daily Presidential Tracking Poll results are reported on a three-day rolling average basis. So, results released on any day of the week were collected over the preceding three days." Thus, it seems they are just averaging their three-day average as it was reported on various days. They are not really providing any new information (everything they present can be computed from the numbers they have already published)...if they had reported the averages of the results obtained (not reported) on the various days of the week, then it would be a different story. In any case, good idea...maybe Gallup would be more amenable to this type of analysis.
Naomi--
McCain is rising in the nat'l polls and Obama is falling; an 8 pt gap was closed in four days. No evidence that the trend has been arrested.
Jinnatonix--OC I am woried about McCain's chances. I'd give the elction to Obama today, if I had to bet. Not enuf empirical evidence to suggest otherwise, but I like the trendlines and the way the narrative is developing.
McCain made an excellent, hard hitting speech at the Nat'l Urban League today. He'll get more favorable publicity tonight and he is not letting up on Obama.
Fabian: it would be a mistake to count the Republicans out. A foot has been dipped in the river and it has been changed forever.
Still developing . . . !
I agree with some of the other posts that comparing a poll with an exit poll isn't terribly useful. What we really need to do is compare Obama's numbers with Kerry's poll numbers the day before the election.
I don't have it broken down the way you do, but I do have some of it by state. All Obama/McCain numbers are 538 projections.
http://www.realclearpolitics.com/bush_vs_kerry_sbys.html#ar
FLORIDA: Bush +0.6 McCain +2.1 R1.5
OHIO: Bush 2.1 Obama +1.6 D 3.7
PENNSYLVANIA: Kerry +0.9 Obama 6.8 D5.9
WISCONSIN: Bush +0.9 Obama 8.8 D9.7
IOWA: Bush 0.3 Obama 6.5 D6.8
MINNESOTA: Kerry +3.2 Obama 8.8. D5.6
MICHIGAN: Kerry +3.5 Obama 3.3 R0.2
MISSOURI: Bush 4.2 McCain 1.6 D2.6
NEW MEXICO: Bush +1.4 Obama 5.4 D6.8
NEVADA: Bush +6.3 McCain .9 D 5.4
COLORADO: Bush +5.2 Obama +2.0 D7.2
NEW HAMPSHIRE: Kerry +1.0 Obama 4.1 D3.1
MAINE: Kerry +9.5 Obama 13.1 D 3.6
WEST VIRGINIA: Bush +8.5 McCain 5.6 D 2.9
OREGON: Kerry +4.8 Obama 8.8 D 4.0
NEW JERSEY: Kerry +7.0 Obama 9.6 D 2.6
ARKANSAS: Bush +6.5 McCain +12.7, R7.2
HAWAII: Bush +0.9 Obama 19.9 D 19.0
An 8 year old could put together a campaign plan for Obama. Your entire campaign should be how McCain is out of toutch with middle america on the economy and he believes the recession in all mental.
There you go, thank you, election won. Just use their own words against them and how the republicans policies and the BUSH/MCCAIN policies are disastrous, see the last 8 years.
McCain quote "The economy is in fundamental good shape." Yeah its in good shape if your Exxon Mobile or a cooperation. The average wage has gone down this last eight years while the cooperations have been laughing all the way to the bank. Thats your entire campaign. "How can McBush help you out if he believes the economy is fundamentally good." he can't.
HIS CHIEF ECONOMIC ADVISER BELIEVES THE ECONOMY IS FUNDAMENTALLY STRONG
How can the economy be fundamentally strong if average wages have gone down. Well, cooperate prophets have been great! So the economy is great.
Its that easy. push McCain as out of touch the same way Bush is every minute of every day. Just attach a catchy slogan on it like Clinton did with "It's the economy stupid!"
The big surprises to me in the differences is that Obama's beating Kerry's final poling numbers in West Virginia and New Jersey.
For those advocating Hillary as VP in order to give the Obama campaign teeth, you are missing the fact that Obama is in charge. It's not like Obama doesn't have people who can be attack dogs. It's that he doesn't want to turn them loose. A HRC VP would be a restrained Hillary; so you would get all her negatives (and write-off a few avenues to 270) but none of her offense.
Scott: I take your analysis. I think the singles number is a construct of the youth number and the AA number (higher divorce rates I believe in this community).
GregM: Oh how ambarrassing. But curious too, since obviously for the kind of analysis Rasmussen was presenting the 3-day averages are less useful. I think he must have foreseen this kind of use and avoided publishing single day based averages for this reason. The tracking polls are not merely being helpful, Scott: I think they also take a proprietary attitude towards their raw numbers, presumably because the rawer the numbers they reveal, the more they also reveal implicitly about their weighting formulae and other tricks of the trade.
Isn't it a little unfair to compare Republican and Democrat primary turnout after Super Tuesday? I mean McCain had the thing locked up in February, while with the Democrats it was a WWI Western Front-style bloodbath until the Saturday after the last primary!
So, isn't that just a little misleading?
"McCain is rising in the nat'l polls and Obama is falling; an 8 pt gap was closed in four days. No evidence that the trend has been arrested."
Another false statement from "Pete Kent" as the Gallup poll has showed expansion and contraction between Obama and McCain before. Check your facts, please.
The press is running from the McCain/Rove mudslinging and it is costing McCain his only asset: his "maverick" brand. Maybe Karl Rove can reinvent McCain as the Mudslinging Maverick.
McCain: George W. Bush's economics and Karl Rove's politics.
VC-
Democratic primary turnout was clobbering GOP primary turnout on and before Super Tuesday as well. It wasn't really a new development. About the only thing super Tuesday changed was McCain's share as the voters (begrudgingly) rallied to him.
Here's my thought about Obama's campaign strategy. I'm not saying I agree with it, but here's what I think they're up to.
1. They're very confident that they can turn out enough people to win in a small number of key states that put them over 269 electoral votes (Bush's winning strategy in 2000 and 2004). This is their fallback strategy. Let me emphasize this: this is their firewall, the thing that wins them the election no matter what. Given the convention siting, I'm betting on Colorado and New Mexico. Anyone betting on Richardson or someone from Colorado for VP? For that matter Kansas is next door to Colorado, don't forget.
2. Their main strategy is to blow McCain out in a landside, with a very Reaganesque positivism. Right now they have the advantage of being ahead in the polls and in the enthusiasm gap. So long as those two things remain true, they will hold to this strategy, with the aim of giving a huge convention speech. I'm betting that Hillary Clinton either introduces Obama or his VP choice. Bill gives a rousing speech early on, too, as does Carter. Rile up the democratic faithful, but also, with Obama's speech, aim to continue to occupy the middle ground and even suck up Republicans. Remember Bill Clinton's huge winning strategy? Welfare reform. With that, he won over enough centrists to ensure a huge reelection victory.
If you look at the data that underlie Nate's post above, you see a very interesting dynamic in the last week in the "Red-Blue-Purple State" dynamic. I'm a little surprised no one's pointed it out. The overseas trip had a huge impact, but perhaps in a surprising way. Obama's Red State support increased dramatically. At the same time, his Purple State support dropped by a similar amount. Thus, although the overall numbers didn't change, Obama is now +2 in Purple States and -3 in Red States.
So, what does this mean? If it's real (and there hasn't been much variability in these numbers in prior weeks), and it was on purpose, it means that Obama is seeking to win Red and Purple States, in other words, to go for a landslide. More power to him if he can pull it off.
3. As for attacking McCain, why do it now? Why not save his attacks for September and October, if necessary? I think this is part of the wait and see approach to #2 above. If he doesn't need to attack, why should he? He only degrades himself. But, if he needs to attack, he can do it if he wants to in the fall. He knows that he can set the tone of the campaign after his convention speech and also that there will be a month less time between his convention and the election than there was in Kerry's case. Remember, the Swift Boat attacks came in August last year, in between the Democratic and Republican conventions. There will be no time for that to happen this year, as the two are right on top of each other. The time between the conventions will be dominated by Obama the Rock Star and McCain's choice of VP.
The game starts after the Olympics folks. If Obama's strategy works, he wins in a landside. If it doesn't, he wins the same way he did in the primaries, by knowing what the exact minimum is that he needs to win.
Comparing the 2004 November exit poll against 3 weeks in July 2008 - during which Obama had his best Gallup polling to date - makes zero sense.
Nate should be tearing this nonsense apart, not celebrating it. Oh, but of course, the analysis was positive for Obama, so it must be right. Why else would Nate so harshly criticize Steven Stark for suggesting Obama may have some hurdles to overcome and then praise this pro-Obama statistical obscenity?
This is not a bi-partisan web site. It never claimed to be. Nate is Pro Obama. Obama inspires him to use his skills and do all this. So stop your crying and go to Real Clear Politics or Rasmussen if you want to hear right wing conversation.
Nate has no obligation to be fair in his commentary or the subject of his blogs. As long as the numbers are accurate and fair, he can talk about Obama all he wants.
"Alex S. said...
Obama almost cannot attack, he chose and is forced to rely on defense. He runs with the "change"-brand. And that resonates with a lot of people. His attacks might be valid, yet the McCain camp will surely find a way to re-interpret these attacks as "Obama going the low road, he is just like other politicians". And there might be a critical mass of people who had enough of the old-style-politics."
People forget WHY Bush's Swift-Boat attacks worked in 2004. But, here's another poster who gets it!
It worked in 2004 because there were more Republicans than Democrats. It won't work in 2008 because there are more Democrats than Republicans. That simple.
Obama doesn't have to sling mud with McCain. He has the +9.5% party ID advantage! McCain doesn't.
All Obama has to do is defend himself. Then McCain's attacks will simply rally the base.
He's doing that right now with ads airing in all battleground states: http://www.youtube.com/watch?v=zPPLSHKH0h4
I saw this yesterday and right after it CBS local news came on with a "Fact-Checker" segment that concluded that McCain's attack on Obama for "not visiting the troops" was false.
So, McCain's even starting to get some push-back from the local media.
Bush polarized the race in 2004. That worked to his advantage because Rove organized 11 million MORE Bush voters than in 2004. Kerry got 8 million, but he couldn't keep up with Rove's GOTV efforts.
And it was the negative ads that drove those voters to the polls!
But McCain is ALREADY at 88-89% of Republicans. There's very little room for him to improve on this! He needs Indpendents BIG time.
If all he does is polarize the race and convince every Republican to turn out and vote, he loses! Because at the same time, he's infuriating Democrats and convincing them to vote for Obama!
Just like Bush did in 2004. Only he didn't have to care, because he had more Republicans. McCain doesn't.
I thought the electorate was 50/50 in 2004 and 2000, and Bush convinced almost all Republicans and the bare minimum of independents.
McCain needs Republicans+PUMAS/right leaning Democrats and a bar majority of independents and he can do something similar.
I just wonder how the PUMAs will react to negatively going after Obama.
NC Moderate said:
"Maybe Karl Rove can reinvent McCain as the Mudslinging Maverick."
LOL
Pete Kent said,
"McCain is rising in the nat'l polls and Obama is falling; an 8 pt gap was closed in four days. No evidence that the trend has been arrested."
No evidence except the entire campaign, which has shown McCain to have a ceiling of 45%-46% of the vote. McCain is going to have to figure out a way of attracting voters to his side, not just whittling down Obama's numbers. Negative campaiging by itself isn't going to get him where he wants to go.
Actually, what McCain really needs is some sort of mind control device that will erase everyone's memory of GW Bush. Without that, I don't see how he wins.
Clark, I agree that that is more or less the Obama campaign strategy.
Kerry was really screwed over by the early timing of the convention (a deliberate choice made to maximize the amount of time the candidate would have access to federal funds).
The convention was at the end of July; there were then two weeks until the Olympics opened on Aug 13. They ran until the 29th, with the Republican convention immediately following.
The Republicans made optimal use of this timing, with the swiftboating killing Kerry's convention momentum beginning on August 5. The combination of the Olympics and Kerry needing to conserve public funding money for the fall made it hard for him to respond right away. Then the Republicans got their convention bounce, and were able to ride it all the way to November (although just barely, and Kerry was gaining in October).
The timing this year is less advantageous for Republicans to smear Obama. The Olympics start before either convention, and it will be difficult for a negative ad message to really catch hold (not to mention Obama has plenty of cash, and is already making huge Olympic ad buys, so he can respond to attacks much faster than Kerry could). August is thus not as propitious a time for smearing as it was in 2004. Neither is the week after the Democratic convention -- that's the week of the Republican convention, and it would be bad strategy to have the Republican convention overshadowed by smear attacks against Obama. Then we're into September, and events like the debates have more influence.
So, at least, it seems to me right now. We'll see how right or wrong I am!
The thing that right-wingers seem to be UNABLE to understand is that Democrats and Independents see these McCain attack ads and HATE them!
Republicans see them and LOVE them! They think "Great! Finally McCain is taking off the gloves!"
That's because they are ALREADY hating Obama and wanting him to lose. So, anything that savages Obama is great!
But, that's not how undecided voters see things. To Independents these are vicious negative ads. They might be true or false, but they are nasty and unseemly. They turn people off.
They also drive up Obama's negatives, because there's this residual effect of "maybe the ads are true!" This encourages McCain to make MORE attacks in the delusion that this is working.
But, all he's doing is polarizing the race. And in the end he can't win a large majority of Independents by this strategy. In fact, the more negative the race gets, the more Independents get disgusted with both parties and vote for a 3rd party or stay home.
Because the key demographic in this race is DEMOCRATS who aren't yet sold on Obama, this only HURTS McCain! It drives Democrats home to Obama because of the unfair nature of the attacks!
Just because right-wingers love these attack ads doesn't mean they are a good idea for McCain!
McCain is going to have to figure out a way of attracting voters to his side, not just whittling down Obama's numbers.
When third party candidates are added to polls, virtually all of the support they get comes from McCain's numbers.
"Virginia Conservative said...
I thought the electorate was 50/50 in 2004 and 2000, and Bush convinced almost all Republicans and the bare minimum of independents."
There's NO EXCUSE for this kind of IGNORANCE if you're going to post as much as you do! Go look at Wikipedia: 2004 U.S. Presidential Election.
Bush won by 2.8% because he got 11 million new voters more than in 2000. Kerry Got 8 million more than Gore, but came up nearly 3 million votes short. Bush won Indeps. by 1%, essentially neutralizing them.
The Partisan ID of 37% D - 37% R - 26% I was taken from the Rasmussen tracking right up to the election.
The difference was Rove's GOTV.
But, this year, Rasmussen is tracking Dem Party ID at +9.5%. He also believes that party ID is basically STATIC, like loyalty to a sports team, that only changes slowly over time, and NOT during an election season.
Basically, Republicans are stuck with about a 7% -10% disadvantage during this election.
There's NO way any GOTV effort can overcome that, even if Obama wasn't threatening to have a BETTER GOTV effort than McCain does. That means McCain MUST run as a "Maverick" "New Style Politician." And NOT as a "Typical Bush-Style Republican."
If he runs the same campaign as Bush he'll be CRUSHED by the Dem. party ID advantage.
the more Independents get disgusted with both parties and vote for a 3rd party or stay home.
And McCain's them to win a lot more than Obama does.
You see, I am seeing no evidence that Obama knows how to handle this. He doesn't accuse the McCain camp of anything.
Your perception is remarkably selective. McCain is damaging his brand and undermining his "base" -- the MSM -- by taking the low road, and Obama is doing just what he needs to by repeatedly calling him out on it -- e.g., his "is this the best they've got?" mockery and the ad that quotes factcheck et. al. saying that McCain's claims are "false", "baloney", etc. You now see John King asking McCain if his tactics are legitimate, and McCain giving a mumbling and confused response and then walking away. That's the current media story on McCain -- is he running a sleazy campaign contrary to what he said he would do. Even Andrea Mitchell pretty much called the McCain campaign liars, and then they ended up contradicting themselves. The bottom line is that McCain's attacks are backfiring on him.
I think y'all are over-playing your hand with party ID. But its a risky strategy that rests on some shaky assumptions, namely that
1) The PUMAs will come home for Obama
and
2) Conservative Democrats will vote for him
Thats who these negative ads are aimed at I think. Its where the "race card" thing was going too. To get the PUMAs who thought Hillary was unfairly accused of being a racist to think twice about Obama.
Darren:
At what point has Nate claimed that he doesn't have a horse in this race? You act surprised that 1) he favors Obama and 2) it is reflected in his posts. If you don't like the pro-Obama tilt of the posts, then don't read them.
The accuracy of the projection itself can only be confirmed on election day, but right now Nate seems to be hanging pretty closely with other sites:
Pollster.com: Obama 284 (without toss-ups)
Coldheartedtruth: Obama 293
Federal Review: Obama 296
538.com: Obama 307.1
EV.com: Obama 316
RealClear Politics: Obama 322
Election Projection: Obama 325
It seems your only argument is that Nate is a man with an opinion - an opinion apparently differing from your own.
If it doesn't, he wins the same way he did in the primaries, by knowing what the exact minimum is that he needs to win.
It's worth keeping in mind that the Obama campaign predicted virtually every outcome, with uncannily good accuracy, except for Maine.
I think Obama should go after McCain's core brand. McCain casts himself as a straight-talker, a man of integrity, and someone who's willing to break with his party on issues when he feels it's correct.
Obama has done a good job of attacking that last third, tying McCain to Bush policy-wise. Now McCain is risking the first two-thirds of his brand by going negative. And Obama has to punish him for it, not just trust that the MSM will do its job or the voters will see through the mud.
"McCain claims to be a straight-talker, but he sure is eager to distort the truth in order to try and smear me. Let's take a look at the evidence...."
"You know, I really liked the McCain of 2000. He ran a hard but clean race, and he treated his opponents and the voters with respect. What happened? Now it's, 'Say anything to get elected'".
I don't know if Obama himself should be saying these things, but they need to be the surrogates' talking points. If McCain wants to define himself downward, by all means let's help him out.
About campaign tactics, the Obama campaign has sent a video to supporters saying they had 200,000 contributors in the last week, and 100,000 yesterday alone, 1/3rd first time contributors. So McCain's negativity has some positives for Obama.
Last day of the month. Expect another huge day of contributions on 4th as that´s his birthday.
Don't want to be just a dittohead, but ...
lilinev (3:30), I think you're right on the money. All Obama has to do is get people to look at the evidence and let them draw their own conclusions about what McCain is doing with his ads.
MATT J H said:
"This is not a bi-partisan web site. It never claimed to be. Nate is Pro Obama. Obama inspires him to use his skills and do all this. So stop your crying and go to Real Clear Politics or Rasmussen if you want to hear right wing conversation.
Nate has no obligation to be fair in his commentary or the subject of his blogs. As long as the numbers are accurate and fair, he can talk about Obama all he wants."
I think you misunderstand the basis of the complaining by people on here. No one said Nate had to be fair in his commentary or be perfectly neutral and unbiased to put out an accurate reflection of the electorate.
However, when you open yourself up in showing a partisan bias (pro-Obama, in this case) and have the audacity to put up a headline to the side of your site's name that says, "Electoral Projections Done Right," (especially when this site is largely untested in the long run - i.e. over several election cycles or years), you are opening yourself up to (justified) criticism.
From what I've understood and gleaned by following this site for quite a while now, it fared well during the primary season in many instances but proved fairly inaccurate in others.
Anyway, Nate has done a good and thorough job of compiling and analyzing the statistics, but putting together a seemingly accurate set of polling data to predict the results in no way obligates the rest of us to cower down and bow to his almighty methods.
As I said, this site and the predictions contained within are largely untested. We can do all of the partisan bickering and mudslinging we want till we're blue in the face and make claims of "accurate and fair" on one side and "biased and unfair" on the other, but until we see how this whole thing plays out in a little over three months, absolutely NO ONE on either side has really any right to talk shit about anything they perceive, believe, glean, feel, suppose, etc. about the data.
Only one poll matters in the end, the one on November 4th.
If Nate's model does a pretty good job then, I'll tip my cap and see that all the claims were true, as all supporters and detractors alike should.
If he's wrong, then I think a lot of people will look rather foolish for their arrogance so early on about an untested model.
VC-
The assumption is that PUMAs aren't going to vote for McCain in numbers that matter. Even assuming that they exist, they can broadly be broken down into two groups: The "Sexism is more prevalent than Racism"-crowd and Reagan Democrats. The latter group doesn't matter. They only make up 19% of the party and only 30% of them are currently saying that they're going to vote McCain. To get them to swing his way, McCain is going to have to do more than campaign negatively. He'll actually have to give them a reason to vote for him, and thus far, all he can talk about is Obama. As for the other group of PUMAS, they might not vote for Obama, but there are other horses in the race. The "Sexism is more prevalent than Racism"-folks aren't going to vote for John "I called my wife a whore" McCain. That's probably why we're currently seeing Obama's lead increase when people are given the choice of voting for Barr or the Green.
MR"However, when you open yourself up in showing a partisan bias (pro-Obama, in this case) and have the audacity to put up a headline to the side of your site's name that says, "Electoral Projections Done Right," (especially when this site is largely untested in the long run - i.e. over several election cycles or years), you are opening yourself up to (justified) criticism. "
The critiscism is not justified when Nate's regressions are acting as leading indicators for state polls. He is "doing it right", but the right doesn't like the results. TS. Get a better candidate.
The Gallup data trove does show that Obama has problem with conservative Democrats that has improved by only 5 points since mid-June. But you also find that McCain never had a problem with conservative voters within his party. McCain's internal problems are actually with liberal to moderate Republicans where he has lost some ground.
The Republican's major weakness is that his electoral support is very fragile. Not only are Republicans less enthusiastic about their standard-bearer but he is more subject to a protest vote. When pollsters add 3rd and 4th party candidates to their questionnaires, McCain is no longer competitive.
Just a heads up, with all these negative britney, hilton/moses, attacks the Obama Campaign received 100,000 donations in the last 24 hours ahead of the reporting deadline.
Approximtely 30% of which were new donations.
Keep it up McCain. This might not scare you if this were going to be a TV campaign. But with a ground game like Obama's, this kind of money can go a long way.
"Virginia Conservative said...
I think y'all are over-playing your hand with party ID. But its a risky strategy that rests on some shaky assumptions, namely that
1) The PUMAs will come home for Obama
and
2) Conservative Democrats will vote for him
Thats who these negative ads are aimed at I think. Its where the "race card" thing was going too. To get the PUMAs who thought Hillary was unfairly accused of being a racist to think twice about Obama.
You STILL don't get it! Bush didn't "peel off conservative Democrats" by Swift-Boating Kerry! He excited REPUBLICANS! Right now Rasmussen finds that 30% of Conservative Dems. will vote for McCain!
Only problem! That's exactly 5% of total Dems! There aren't very many "Reagan Democrats" left! They all became Republicans.
The only way for McCain to appeal to either dissatisfied Democrats OR Independents is to appear to be the 2000 McCain!
He would have to run as a "Maverick."
No negative attack ads, a "new style of Republican", acknowledge his opponent, but proclaim his "experience", etc. In short he has to appear to be bi-partisan, not a knife-fighter.
He had a chance to do this, but he concluded that there was NO WAY he was going to win as the "change" candidate in 2008 and he needed to go full-blown NUCLEAR!
Well, that is exactly what Bush did in 2004. But, it only worked then because Bush had a MUCH bigger base than Republicans do now.
You righties never get it correct! You seem to think this "partisan ID nonsense" is something LIBERALS made up to make Obama look good! You think a few attack ads by McCain are going to change it!
Rasmussen and other pollsters have been measuring this for years and all the signs point to a HUGE partisan ID advantage in this election. That's fixed in stone and it's not going to change.
All you have to do is ask yourself: is the way to appeal to any disaffected Democrat to savagely attack the Democratic nominee?
No!
I've been following this site since the Primaries and this is my first comment. I see that there is some debate here on Nate's objectivity, given his support for Obama (which support I share).
I don't think we have to worry about Nate's objectivity since he has a reputation from his sports work as well as a reputation that he is trying to build in this arena. In the end, he will be true to the data.
On November 4th, there will be a public reckoning for him: he will either be proven right or wrong. Personally, I've been working in or around the political/electoral realm for many years and I like to kid myself that I might know a thing or two about stuhtisticks. I've been tracking Nate's projections (top line, not state by state) on a daily basis since the beginning of July (I gave him a few weeks to work the bugs out of his model).
He has a very interesting methodology, he has the guts to put it out there for us to see every day and he certainly seems to bring statistical bona fides to the table. It's going to be an interesting run. Let's enjoy it and have a lively dialogue along the way.
My own view, on which I will be happy to expand in the coming weeks, is that it's too early to make much of the polls, but that the numbers don't look as good as they should for Obama. But, this is long enough for an initial comment.
Nate clearly knows what he's doing. Together, we'll find out whether he's right and, hopefully, have some fun along the way.
Mason,
I see your point. You didn't say anything I didn't allude to...yes, Nate is doing it right, mathematically speaking. And, yes, to some extent his model may be serving as a valid indicator.
But as ignorant as some have been on the right to downplay the model over perceived bias or lack of statistical understanding, there are as many who blindly and arrogantly defend it as "right," when it technically hasn't predicted squat yet. This is August 1st. The election is November 4th.
Don't get preachy to me about "doing it right" with "statistical indicators." I have a master's in economics, so I know a thing or two about using stats and models to prove or disprove something. A good model can be a great tool and excellent indicator for whatever you're forecasting, but I've seen the best fail at any given time under a multitude of circumstances.
So keep on being the object of my admonition and talk shit all you want. On November 5th, you'll either be able to raise more of a ruckus in triumph or you will look very foolish (assuming the model changes little between now and the election).
And, by the way, don't talk down to me and tell me "Tough shit, get another candidate." I'm a conservative independent observer. I have no candidate. McCain or Obama. Or Barr or any third party either. The political system has sold this country out, so I keep my head down, mind my own business, and work hard and let all the yahoos sit around and worry themselves an ulcer about who's "in" or "out" of office and how that makes all the difference in the world.
NC Moderate:
Where do you come off saying my statement about McCain closing the gap in the Gal tracker is false? Your point about there having been variation before is a non sequitor. You do not reason well.
Naomi: You are so snide and smug. But will you be willing to eat crow if McCain does break through this supposed ceiling?
The first step in candidates reversing position is the undecided vote goes up. If that number persists and the primaries are any guide, the lion's share of those votes will go to McCain.
And Cugel: What do you know about independents? You are a Daily Kos wannabe from day one! Stop pretending you know about a group of people you have never been a part of.
*******
Nancy Pelosi shut down the House today in a brazen move worthy of the Soviet Politburo. The stubborn resistance of the Democratic leadership and their standard bearer, Barack Obama, to refuse to allow drilling will hurt them. Strong arm tactics such as these will be viewed in a harsh negative light.
Expect a bumpy road for Obama and the Dems over the weekend.
In order to feel some confidence in their chances, the Republicans need to be aiming to be very close to winning at least two large states that went to the Democrats last time in order to compensate for probably losing Iowa, New Mexico & Colorado. There is one such state: Michigan. But the other state they have a shot in - New Hampshire - is small. That means things are tough for McCain.
Bravo, JohnNYC...
Now I can appreciate a good comment made in humility, regardless of who they support.
You hit the nail on the head with your observation and more eloquently portrayed some of the same things I was trying to say.
Here's to enjoying watching this thing play out over the next several months, regardless of the outcome.
McCain is a one issue man - drilling. Why not take it away and say, well, lets leave it to the states.
McCain is a one issue man - drilling. Why not take it away and say, well, lets leave it to the states.
Pete,
you're being equally presumptuous in assuming that this decline in the tracking polls is going to be any different from any of the others. Some people see these results and think it's the beginning of the end for Obama; others see it as essentially random fluctuation of a kind we've seen before in this campaign.
If and when McCain
1) actually leads in more than one in 50 polls,
2) begins to lead consistently in the polls,
3) begins to have a 2+ point lead in the polls,
then I'll begin to believe he's turned the campaign around (and if he were to move from 1) through 2) to 3) I'd take that turning-around seriously).
For the moment, I don't see anything particularly bad for Obama in the numbers, and I also don't see that anything substantive has happened to change the fundamental shape of the campaign so far.
Post a Comment