Presidential Forecast, 11/3: 44 Hours Left

by: Chris Bowers

Mon Nov 03, 2008 at 03:00


Electoral College: Obama 291, McCain 160 Toss-up 87 (270 to win, 269 to tie)
National popular vote: Obama 51.6%--44.6% McCain


Dark Blue (273): Obama +6.0% or more
Lean Blue (18): Obama +2.3%-+5.9%
White / Toss-up (87): Obama +2.2% to McCain +2.2%
Lean Red (28): McCain +2.3%-+5.9%
Dark Red (132): McCain +6.0% or more

Relevant States Chart
Polls 10/28-11/02

States not shown are further either direction
State EV's Obama % McCain % Margin Polls Obama EVs
Obama Lock 154 154
Iowa 7 54.5% 38.5% +16.0% 2 161
Maine-AL 2 54.0% 38.3% +15.7% 3* 163
New Jersey 15 53.5% 38.0% +15.5% 2 178
Oregon 7 56.0% 40.7% +15.3% 3 185
Washington 11 53.5% 39.0% +14.5% 2* 196
Wisconsin 10 54.0% 40.5% +13.5% 2* 206
Michigan 17 53.7% 40.7% +13.0% 3 223
Maine-02 1 52.0% 41.0% +11.0% 1* 224
New Mexico 5 53.8% 43.3% +10.5% 4 229
New Hampshire 4 52.6% 42.4% +10.2% 5 233
Minnesota 10 53.3% 44.0% +9.3% 4 243
Pennsylvania 21 52.0% 44.5% +7.5% 6 264
Colorado 9 51.7% 44.3% +7.4% 3 273
Nevada 5 49.0% 43.0% +6.0% 2 278
Virginia 13 50.0% 45.3% +4.7% 4 291
Florida 27 48.0% 46.0% +2.0% 3 318
Ohio 20 48.3% 46.3% +2.0% 3 338
North Carolina 15 48.2% 48.0% +0.2% 6 353
Missouri 11 47.0% 48.8% -0.8% 4 364
North Dakota 3 46.0% 47.0% -1.0% 1 367
Indiana 11 46.3% 48.3% -2.0% 4 378
Arizona 10 46.5% 49.0% -2.5% 2 388
Montana 3 45.3% 49.0% -3.7% 3 391
Georgia 15 45.5% 49.5% -4.0% 2 406
Lock McCain 132 538
* = One or more polls in average conducted either partially or entirely before October 28th

Analysis and methodology in the extended entry.

Chris Bowers :: Presidential Forecast, 11/3: 44 Hours Left
Methodology
All of the polls included in the averages had all of their interviews conducted in the final seven days of the campaign (October 28th forward). Campaign-funded and non-random Internet polls are not included. Every pollster only has one poll per average, and all polls are weighted equally. In the event two polls were not taken in the final seven days of the campaign, the two most recent polls were used instead. No polls conducted before October 13th are ever used, no matter what. The states listed were either swing states in 2004, or are swing states now. My rationale for this methodology is based on research I conducted back in June, which you can read here.

Analysis
After brief periods of improvement, Florida and Ohio once again fall well behind the other swing states in terms of their shift from 2004. This might have something to do with the Mason-Dixon polls, which do indeed tilt Republican, but it might also represent real tightening in those two states. It doesn't effect the outcome, since Obama is a lock in Colorado and Pennsylvania, but it would make Obama's victory a bit less impressive and a bit less satisfying. In a rare moment of optimism, right now I am prone the believe that it is just a "house" polling effect.

Except in the few states--indicated by an *--that are still using older polls, no more polls will be removed from the forecast, only added. (Polls can be replaced if one polling organization releases a new survey that supercedes a previous one.) The final numbers are hardening, and there just doesn't seem to be any way that McCain catch up in Colorado, Iowa, New Mexico, or a single Kerry state (or Nevada and Virginia, too). Obama has enough to win, and even if it might not be as huge as some had--or still--hope for, it still appears to be in the bag.


Tags: , , , , (All Tags)
Print Friendly View Send As Email
Chris (4.00 / 2)
you  missed a slew of Zogby polls for a lot of these states released tonight.

They will improve Obama's average in Pennsylvania, Missouri, Ohio, Virginia, Nevada.

About even to the trend in Florida.

Here are the numbers:

PA
O: 53.7
M: 40.0
OH
O: 50.2
M: 43.9

VA
O: 50.7
M: 44.6

NV
O: 50.7
M: 42.9

FL
O: 47.5
M: 46.2

MO
O: 47.4
M: 45.7

NC
O: 47.7
M: 49.3

IN
O: 43.9
M: 49.1

They'll hurt him Indiana. Check Zogby's website.


Colorado (0.00 / 0)
Chris,

Could you elaborate on CO being a lock? I would sleep a lot easier.


There's been a shortage... (0.00 / 0)
...of high quality polling in CO recently.  However, what there has been has never shown Obama with less than a four-point lead (even Mason-Dixon, which has come out with much closer spreads in a slew of states than other reputable pollsters, had it 49-44 as of yesterday).  Combined with the tremendous early voting advantage Obama's built up, it's very unlikely Obama will lose CO.

FiveThirtyEight.com puts Obama's win chances at 96% for Colorado.


[ Parent ]
Oh, another poll (0.00 / 0)
It's not quite as recent, but PPP had it at 54-44 Obama four days ago.

[ Parent ]
How do some pundits arrive at "odd" numbers? (0.00 / 0)
Look at the WaPo crystal ball contest:
http://www.washingtonpost.com/...

Well, some strange numbers there, for instance Chris Cillizza: "Obama 312 : 52%McCain 226 : 48%"

312??? Looking at your table, 313 would make some sense (Obama winning Va and Fl, but not Nevada). But 312? Do the states that allocate their electors proportionally to the vote result in the math becoming a bit fuzzy? And, if yes, how many electors are likely in the play? One or two, or even more?  


Maybe he thinks McCain will win ME CD2? (4.00 / 1)
Remember, Maine and Nebraska aren't winner-take-all.

[ Parent ]
Sweep... (0.00 / 0)
A sweep of all those states listed looks mighty nice...  406 EVs.... hmmmm.

Hopefully the R's are so depressed they don't vote Tuesday. =)


Funny how the original "firewall" (4.00 / 3)
Is the one that lasted... It looked like VA was going to replace CO as the most solid firewall for awhile... but no.. back to the whole Kerry + IA, NM and CO deal. =)

Simulated Election Night (0.00 / 0)
Has anyone done a projection based upon poll closing times to show how the EV's will come in on Election Night and determine which state should put BO over the top?

hmm, that would be nice (0.00 / 0)
All polls close at 9pm, right?  It would be good to see this chart in order of time zones...  

For myself and a friend I'm watching with, I'm going to keep one of the interactive maps open, with all the swing states from Chris's chart unassigned at the outset.  And click away as the states are called.  Maybe a slug of my beer for every blue state. :-)


[ Parent ]
on second thought (0.00 / 0)
I think I have to give Obama a few more blue states at the start: NJ, WA, OR, WI & MI

I don't really want to sweat OR and WA until the polls close on the west coast!


[ Parent ]
Staggered closing times (0.00 / 0)
Not all polls close at 9 pm. Not at all. Remember that voter suppression has long been a tool of the ruling party.

Polls close in the eastern, larger parts of Indiana and Kentucky at 6 pm. That helps keep those unionized shift workers from coming in after work and messing things up! And those early closing precincts will not be released until the others close at 7 pm EST, when the networks will call KY and probably call IN, based on exit polls and a handful of sample precincts fed into their computers. If they say, Indiana is "still too close to call," at 7 pm EST, start dancing, because it means we will win the election no doubt about it.

See Swing State Project for a map of the states colored by their closing times.


[ Parent ]
Closing times (0.00 / 0)
Swing State Project has a map with closing times for each state as well as an accompanying chart listing the hot races in each state.  Indiana and Kentucky have the earliest closing times at 6 PM eastern but it only covers part of those states.  The last polls close at 1 AM (eastern time) in the Aleutians and on St. Lawrence Island.  No more than 2,636 registered voters live in those locations.  Despite being in Alaska, these areas have a slight Democratic lean when it comes to registrations.

A national holiday with uniform poll hours (8 to eastern would mean 5 to 5 Pacific and 4 to 4 for 95% of Alaska) would be a great idea.  Unfortunately, holding the election on Saturday or Sunday would disenfranchise some voters along religious lines so it is probably a bad idea.


[ Parent ]
thanks David & Woody (0.00 / 0)
I made myself a little list of states in order of polls closing times, using Chris's chart from 5 days ago.  I left off WA, OR, NJ, MN, WI.  If they aren't safely blue, I'm hibernating for 4 years.

(trying not to feel stupid about my 9pm error :-)


[ Parent ]
if we can't destroy the (0.00 / 0)
sod (scourge of democracy, gop) then during the next four years while obama governs showing great respect and deference for the opposition and their followers, and while the right are carping, chirping, sniping, and attacking the entire time, then we will put them out of their misery in 2012, let them suffer awhile it will give them some humility, or maybe not!

Kudos to you... (4.00 / 1)
...for a small-c conservative methodology.

Of course, the unallocated ev's have to go somewhere, and while the end-of-race tightening (undecideds have to decide, after all) makes it look like Obama's less likely to win many of the states, I think both Obama being above 50% in a number of these states, such as Florida and Ohio, makes the situation a little more rosy for Obama.

Now, here's what makes me think we are still looking at a blowout: those early voting numbers.

Traditionally, the Republicans either tie that up or better the Dems.  But this year, Obama's kicking butt in early voting numbers pretty much everywhere you look.  What that tells me is simple: either Obama supporters are vastly more enthusiastic this year, or Obama's GOTV/ground game efforts are everything they're cracked up to be.  Or both.

Either way, if I were a McCain supporter, I'd find those numbers to be an extremely ominous "canary in the coal mine" regarding voter behavior on election day and the final vote tallies across the board.


Latest NBC/WSJ poll (4.00 / 1)
Conducted entirely on Sunday puts it at 51/43 Obama.  Allocating undecideds evenly gives Obama an eight point victory.

Assigning perhaps two points to third party reduces that to six.

I'll take it.


Wow, both you and Nate Silver posted 3 a.m. updates (0.00 / 0)
You guys aren't going to get any sleep until Wednesday, are you?

Could be a scheduled posting, executed automatically... (4.00 / 1)
in order to provide a steady stream of new stories for the blog. Quite common, this technique, Atrios does that regularly, too.

However, Chris posted his last comment at 1:34 am, so he really stayed up late!


[ Parent ]
Ohio trending Obama this morning (4.00 / 1)
Q-poll +7, U. of Cin. + 6, Zogby +6, Columbus Dispatch +6

Put Ohio in the win column.


Yeah, those new Ohio number have me very relieved this morning (4.00 / 1)
[ Parent ]
did you do separate breakdowns for nebraska's votes? (0.00 / 0)
if so, can you give a sense of where they're at?

he does, so they're worse than 4% off (4.00 / 1)
The state as a whole is 58-35 according to pollster.com, so we're does 23.  Obama does something like 11 better in Chris's projections in the best NE district, which still leaves it off the chart.

New Jersey politics at Blue Jersey.

[ Parent ]
USER MENU

Open Left Campaigns

SEARCH

   

Advanced Search

QUICK HITS
STATE BLOGS
Powered by: SoapBlox