Talk Elections

SUSA says they're going to release GE polling for all 50 states today:

link (http://www.surveyusa.com/index.php/2008/03/06/new-electoral-maps-to-be-released-today-based-on-30000-just-completed-interviews/)

They're already revealing who they have leading in every state.  Obama v. McCain:

()

Clinton v. McCain:

()

However, they seem to be leaving things ambiguous as to whether every single one of these state-by-state polls is new.  Might they just be compiling some of their polls taken over the last couple of months, and calling it a brand new thing?

North Dakota?

Well, cool anyway.

McCain-Clinton is a 276-262 Clinton victory.

()

McCain-Obama is a 278-260 Obama victory.

()


North Dakota is weird, but I've long had the suspicion that Obama might make the rural northern west a lot more competitive than it has been, tapping into the extraordinary anti-establishment sentiment there. (That's really just the Dakotas, Montana and Wyoming.) Maybe that's ridiculous optimism, but there you go.

Also, the usual caveats about polling certain states apply, NJ being the most obvious on these maps.

Glad to see Obama is leading in Ohio. Trailing in NJ though? Not likely.

Did they just poll some of BRTD's people?

Ohio and Pennsylvania should be flipped on Obama-McCain.

This is exciting but do we not get percentages? 

When the polls are released later today, yes.

The interviews were "just completd," by the way, so we know they're all new polls.

The West likes Obama aww yeah

I would say North Dakota is a tad less believable than New Jersey, but ok. :P

SUSA have been good with primaries but they've never been very good with GE polls. They're an indicator, but their national maps have always been a bit crazy.

Apparently one of the Democrats won a CD in Nebraska. My guess is Obama in NE-02

Eh - we'll see.  North Dakota is less anti-establishment than either South Dakota or Montana, and more Republican.  I'll want to see the polls out of those two states.  Idaho and Wyoming, too.  Even parts of Nebraska; maybe Obama will "win" a CD there, not that the poll would show this*.  Should be a fun afternoon.

But North Dakota is not voting Democratic in the General unless the Dem candidate is named Bartlett and is fictional.

* - Edit: ...or maybe it will.  That was sure a weird coincidence.

But the West loves Obama!

Seriously, I think these polls are sort of correct in the tendencies they show. But SUSA tends to blow up a couple of polls pretty badly when they do this. In the Obama v McCain match-up I'd say Ohio, North Dakota and New Jersey look completely unbelievable. I don't buy Virginia either. In Clinton v McCain the swing seem a little too large (McCain winning New Hampshire and Washington, Clinton winning West Virginia and Florida) but the trend is correct. Apart from Michigan, lol. Michigan going GOP when Ohio, Pennsylvania and Florida don't? Yeah, right.

Niiiiice.

I´m in for a long night to enter this stuff to the database. :)

All in all, these look pretty standard for polls done before a candidate is even finalized.

ND? LMAO

Btw:

Clinton losing MI, NH, WA and OR to McCain ? Possible, but unlikely. Obama leading in ND is probably the biggest outlier here. I don´t believe it until I see other polls from there. Everything else on the Obama vs. McCain map looks credible. I´m looking forward to the LA results and how Obama does compared to Clinton in Arkansas.

()

()

()

()

The exact results were less credible than the map. But MA and TX are beginning to consistently appear to be in play.

Don´t get it. Just a few days ago they released a TX poll showing Clinton closer to McCain than Obama. Now Obama is about to win TX and Clinton is trailing ? :P

Obama down by four in Alaska and South Dakota, three in Nebraska, and one in Texas? Try again guys

SUSA should follow your advice: Poll all 50 states each month now until November. Pollgasm ! ;)

Will wait and see what other polls say over the next few weeks and months... either way Obama vs McCain has the potential to radically recast the electoral map in a way that we havent seen for a very long time.

SUSA probably over states it a bit this far out, but i think the trend towards Clinton/McCain being fought along 2000/04 lines while Obama/McCain being a good deal more open-ended is likley to be reflected in other polls.

I'm encouraged by the Ohio numbers.

Some of these numbers are pretty funny but then you have to expect that (isn't it about one in twenty polls that blows up totally? And here are one hundred polls! And early polls are often a little silly anyways). But the general pattern makes a certain amount of sense.

I didn't say ND was more believable, but NJ has obvious caveats in polling that could explain results whereas ND does not. As usual, in this poll, NJ in Obama-McCain has the highest undecideds in the country, and you know where they'll go.

Even if everything on the maps is correct...I feel REALLY good about FLORIDA...and that helps McCain.

Not that it means much but here is  where Obama and Clinton performed better than each other

()

Ohio is grey because this poll has them both up by 10 over McCain... although it is pretty obvious Hillary would do better there.

No surprises for any of those on where Obama and Clinton do better than each other, except Mississippi and maybe Texas.

^^^^^  Absolutely correct.  (1/20 is right too)

Of course, there's other issues here that are being pointed out, as well as some not being pointed out.  I'll add one in particular that is quite important.  Texas and Ohio numbers should not be trusted if they were in the past two weeks.

What I'm going to try and do later is put the numbers together and weight by state based on turnout the last 4 elections because there's a few things I'm looking for nationally, and 30,000 is quite a good sample.

Looking at the internals, over most states the shift appears generational and racial... gendered too, but to a lesser extent. Clinton underperfoms in under 35's, AA's, and Independents, Obama underperforms in 55+'s, Dems, and Hispanics. 

Does this make anyone else think that a unity ticked would be mutherf***in unstoppable?

One explanation on Mississippi is turnout,  When its all set and done white voters in Mississippi will voter overwhelmingly for McCain no matter who the Dem candidate is, black voters will vote overwelmingly for either Obama or Clinton.  However, turnout among African Americans will be higher wirth Obama than Clinton which would make the state closer, still not enough to win because the white vote will be too strong for McCain.

Do you think Clinton picking Obama would pull in the votes of people on this site (for example)?

Do you think a black man/white woman ticket would pull in the over 55s?

After three months (and possibly more) of this primary continuing, how much infighting between Clinton and Obama supporters is going to cause them to think McCain is a second choice to their first option?

Just a few thoughts to think about, that's all... :)

The people on this site aren't a good cross section of Democratic voters because we're all so obsessed with the race and follow it so closely. The vast, vast majority of voters haven't been following anywhere near as closely as us and thus aren't as radicalized (for lack of a better word) towards their candidate.

The other two points I think are quite valid though.

I suspect the under-35 Obama supporters who support McCain are a bit more radicalized than the normal under-35 population, but maybe I'm wrong.  The analogy was probably not the greatest, I admit.

2 Nebraska electors for Obama? Hilarious.

I doubt anything other than a small portion.  One thing supporters of both Obama and Clinton have in common is they hate bush.  McCain running a campaign on continuing Bush's policies isn't going to endure him to voters who hate Bush's guts.

Obama significantly underperforms in Massachusetts, and this should be no surprise.

After all, his political clone has been stinking things up as Governor here for the past 15 months.

Obama vs. McCain

()

Question 1) This site doesn't exactly represent the polity, now does it

2) Just the Women and the Blacks... so yes

3) People like them both, and there is a general election to overcome gripes.... How many Dean supporters voted for Bush or Nader? If 2 million people have contributed to and volunteered for these campaigns and invested enough to hate the other and stay home. thats what 1% of the possible voter pool?

If Nebraska breaks 45-42 for the GOP, given how NE-3 usually breaks about 10% more GOP than the rest of the State, then the Dem getting NE-1 and NE-2 is believable.

Did anyone else lol when they saw Utah was within 11 points with Obama?

Battleground Utah baby! ;)

Anyone up to running Monte Carlo simulations based on these numbers and give us EC win estimations?

If SUSA's web site actually worked I'd enter them.  I hope they have similar methodologies.

Based on the numbers posted above, I should be able to do one of the three major breakdowns that I'm curious about.  The other two require internals.

National popular vote numbers (averaging last three elections per state and weighting)
National party ID
Candidate % of party ID

I may eventually do a national racial breakdown, if I get to it.

I doubt Obama would be that great in the Mountain/Plain states, but perhaps he will. It would be hard to believe ND voting for Obama, but why not? It is a small state, and the swing of voters isn't that much. I think we could actually see a map like the polls say, somewhat crazy and different, but still plausible.

Wow.  Obama takes two electors in Nebraska.  Carries GOP ND.

Dems, you MUST nominate this man.  The black messiah in unbeatable!

Well, it certainly appears that there are no hard feelings against Obama in FL.

He's not going to win those states but he going to be significantly stronger in those prairie states and every state westward than Hillary.

There are a good number of states I trust SUSA in - with caveats of course.  Florida is not one of them - for some reason they have a pisspoor record there - too much to the Dem side.  But for my purposes, that's not what I'm interested in right now.

All entered.

I think you reversed the numbers in Kansas.

Thanks.  I'll be happy if that's the only error I made.

(Other than entering them half as 2/2 and having to go back... :()

Obama
Kerry states lost: 2
Bush states won: 7

Clinton
Kerry states lost 4
Bush states won: 5

Obama:
Loss: 36 EV
Gain: 64 EV
Net: + 28 EV

Clinton:
Loss: 39 EV
Gain: 63 EV
Net: + 24 EV

Just felt like adding the numerical values.

Obama isn't winning North Dakota.
Obama isn't winning any EV out of Nebraska.
McCain isn't winning New Jersey.
Also on the McCain/Obama matchup: Ohio, Pennsylvania, and Virginia are all opposite the way I think they will go.

Fixed EV Total: Obama 278  McCain 260

McCain isn't winning Washington or Michigan against Clinton, I think Oregon is possible but not likely. Is Clinton really that popular in Florida, Ohio, and West Virginia? I think Iowa and New Mexico should be switched in the Clinton/McCain matchup.

Fixed EV Total: Clinton 313  McCain 225

Nice work.

Wow.  Basically the worst states in that red belt there. ;)

I wasn't aware that suddenly we can be picky about those who vote. ;)

... I do the same when I play P4E - I start leading in a redneck state I stop all ads and visits

From the looks of the internals, Democrats should be praying to God that Barack Obama is the nominee—McCain's relative strength against him in most states is due to defecting Democrats.  Obama's strength is largely due to independents.

Hillary, meanwhile, is solid with Democrats, but performs poorly with independents.

Those Democrats should be far easier to consolidate than the indies.

Also, it's very interesting that Hillary and Obama have different strengths in different states.

If I was a Massachusetts superdelegate, I'd be hard pressed voting for Obama when I know he'll lop 10 points off the generic Democratic vote here.

If I was a Utah superdelegate, no way would I vote for Hillary, considering that Obama actually has significant appeal amongst independent voters there.

It's all about keeping your job, folks.

That presumes, of course, that those are the types of Democrats who can actually be consolidated behind a black man named Barack Hussein Obama.  Without physical force in the voting booth, of course.  I remain unconvinced.

Bump...

I weighted these polling results to an average of voter turnout 1996-2004.  This is not perfect, but it gets out some high/low turnout issues.  I substituted in some reasonable figures for DC that I thought would approximate other things I saw in these polls.

National Popular Vote
Obama 46.87%
McCain 44.07%
Undecided 9.07%

Clinton 45.75%
McCain 44.49%
Undecided 9.76%

Party Affliation
Democrat 44.00%
Republican 31.45%
Independent 19.20%
Other 5.35%

---

Observations
1) The higher undecided figure on the Clinton side is almost entirely due to black people in certain states.
2) I suspect the Obama number is only higher presently because of some of the *interesting* results these polls show in the west.
3) Even though I haven't done the party ID breakdown yet, in 3/4 of the states (minimum), Clinton does better than Obama among Dems (even in states where he performs much better than her), and in nearly every state, Obama does better than Clinton among GOPers and Indys.  Now, since I suspect Democratic affliation is a decent bit higher than we'll see come GE time, that's something to keep in mind.  Also important to keep in mind is whether you think it is easier for Obama to consolidate Dems or McCain to consolidate GOPers and pick off more conservative Indies.
4) If you discard the *tied* states (NJ/VA for Obama; MI/NM/TN for Clinton), you come up with:
Obama 267
McCain 243
Undecided 28

Clinton 271
McCain 234
Undecided 33

Considering Obama performs roughly 1.5% better in terms of the popular vote, this is what I mean when I say that Clinton performs better electorally than Obama does in a close race.

---

Next on the list for me to do is a breakdown of party ID and race and maybe more.

This is consistent with national polls from, for example, Gallup and Pew, which show Obama leading to more Democratic defections to McCain than Clinton would, but Obama still outperforming Clinton in total voters because he attracts more Indies and Republicans than she would:

()

()

I certainly agree - and I had noticed it before too, which is one of the reasons why I did the above work.

I must admit that one of the reasons I prefer Obama to Clinton is that I think the country as a whole is better off with a less polarized electorate. If more Republicans and Independents will support Obama, I feel it's worth it to lose some Democrats. Of course, the accuracy of these polls can (and should) be debated, especially with the small sample sizes of demographic groups.

Of course.  And it is also merely a "snapshot in time".  But that's one of the reasons why I did the national vote numbers...

Because after hard weighting, the MOE of a sample this size is liable to be pretty low, all told.

8 years?

I think the only safe assumption about that is that the share will be smaller come election time than the polls now show.
That mostly means the difference between Clinton and Obama will likely be smaller than the polls now show, too.

One might infer from my first sentence that McCain is bound to do worse on balance than polls now show, but (while possible) I'm not making that claim; Reps are hardly united behind their candidate yet either, and quite a few of them, too, are currently supporting an unusual second choice (Obama or Clinton) after not getting their first choice (one of the non-McCain Republicans). Many of them are going to come back as well. And there's the whole "Dems currently more fired up because the campaign's still going; this shows up in polls" line of argument that I'm not sure I'm buying but cannot dismiss either.

I have no disagreements with your statement - I'm just pointing out the differences.

I was just using the quote as a starting point to say what I was going to say anyhow.:)

"Obama has coattails in Nebraska,'' said Nelson, who endorsed his Senate colleague two months ago. "Our internal polls show he can win one, possibly two, congressional districts.''

http://www.bloomberg.com/apps/news?pid=20601087&sid=a0OkW8Ml8ljw&refer=home