Monday, September 22, 2008

The Electoral College Map (9/22/08)

Sunday's polls offered a mixed bag of results for both candidates. Of the nine polls from 7 states, the pair of surveys out of Florida continued to show a race drawing closer, while following the tie in the Big Ten poll of Iowa, the two polls in the Hawkeye state show the exact opposite. That outlier, then, appears to be just that. An anomaly. Other than those two, both Ohio and Virginia remain well within the margin of error -- despite the wider spread in the University of Cinncinati survey of Ohio -- and the formerly large Obama gap in Minnesota continues to contract. But the race across all the Rust Belt and north midwestern states is heading in much the same direction.

New Polls (Sept. 21)
StatePollMargin
Alabama
Univ. South Alabama
+27
California
ARG
+14
Florida
Miami Herald/St. Pete Times
+2
Florida
Research 2000
+1
Iowa
ARG
+14
Iowa
Research 2000
+7
Minnesota
ARG
+1
Ohio
University of Cinncinati
+6
Virginia
ARG
+2

But Minnesota still has not budged too terribly much on the map or on the Electoral College Spectrum. In fact, the map is unchanged since yesterday's South Carolina shift. Obama maintains his eight electoral vote edge over McCain due in large part to the fact that the five state block from Colorado through Virginia is stationary through the addition of these nine polls. If there is any change to the electoral vote tally, it will be triggered most likely by the movement of one of those five states. And only Nevada and Ohio are currently on the Watch List below.

[Click Map to Enlarge]

Though Minnesota does not shift again, despite yet another close poll, it is within a fraction of a point of jumping Oregon and closer to a toss up distinction. But the North Star state isn't there yet. Of the five polls released since the beginning of September in Minnesota, all have been within three points with the exception of the CNN poll that was conducted in the time between the two conventions. So while the FHQ average is not yet yielding a result in line with this recent trend, it has shifted by almost two points. Four polls, after all, is just more than one-sixth of the number of surveys conducted in the state. We have a trend and the average is following that trend, but cautiously.

The Electoral College Spectrum*
HI-4
(7)**
DE-3
(157)
CO-9***
(269/278)
ND-3
(160)
AR-6
(64)
VT-3
(10)
WA-11
(168)
NH-4***
(273/269)
WV-5
(157)
KS-6
(58)
RI-4
(14)
MN-10
(178)
OH-20
(293/265)
SC-8
(152)
NE-5
(52)
IL-21
(35)
OR-7
(185)
NV-5
(298/245)
AK-3
(144)
TN-11
(47)
CT-7
(42)
NJ-15
(200)
VA-13
(311/240)
TX-34
(141)
KY-8
(36)
MD-10
(52)
IA-7
(207)
IN-11
(322/227)
GA-15
(107)
AL-9
(28)
NY-31
(83)
NM-5
(212)
FL-27
(349/216)
SD-3
(92)
OK-7
(19)
CA-55
(138)
WI-10
(222)
MT-3
(352/189)
AZ-10
(89)
WY-3
(12)
ME-4
(142)
MI-17
(239/316)
MO-11
(363/186)
MS-6
(79)
ID-4
(9)
MA-12
(154)
PA-21
(260/299)
NC-15
(378/175)
LA-9
(73)
UT-5
(5)
*Follow the link for a detailed explanation on how to read the Electoral College Spectrum.
**The numbers in the parentheses refer to the number of electoral votes a candidate would have if he won all the states ranked prior to that state. If, for example, McCain won all the states up to and including Pennsylvania (all Obama's toss up states, but Michigan), he would have 299 electoral votes. Both candidates numbers are only totaled through their rival's toss up states. In those cases, Obama's number is on the left and McCain's is on the right in italics.

***
The line between Colorado and New Hampshire is the where Obama crosses (or McCain would cross) the 270 electoral vote threshold to win the presidential election. That line is referred to as the victory line. Both states are currently favoring Obama, thus the blue text in those two cells.

Finally, the Watch List from yesterday remains intact. These ten states plus the McCain lean states are still the ones to keep our eyes on. Are those medium red states moving toward competitiveness or toward McCain? And will and other states -- such as New Mexico or Wisconsin -- shift to toss ups?

The Watch List*
StateSwitch
Alaskafrom McCain leanto Strong McCain
Delawarefrom Strong Obama
to Obama lean
Georgiafrom Strong McCainto McCain lean
Nevadafrom Toss Up McCainto Toss Up Obama
New Mexicofrom Obama leanto Toss Up Obama
North Carolinafrom Toss Up McCain
to McCain lean
Ohiofrom Toss Up McCain
to Toss Up Obama
Texasfrom Strong McCainto McCain lean
Washingtonfrom Obama lean
to Strong Obama
Wisconsinfrom Obama leanto Toss Up Obama
*Weighted Average within a fraction of a point of changing categories.

Since we have a status quo post here, let me take an opportunity to pose a question to our readers in the hopes of jumpstarting a discussion. I have had several discussions lately about the changing definition of what a toss up state is. In other words, is a state that was a toss up at five points in June or July, still a toss up with the same five point margin today? So the definition of a toss up is something of a moving target, shrinking as the campaign draws toward its culminating point on election day. Now that the system has calmed down to some degree after the convention bounces, I think it may be the best time to move that line, especially with the next series of shocks set to commence on Friday with the first debate. What do you think?

Let's look at a couple of examples, so those wishing to comment can make as informed a decision as possible.

If the line is moved from a five point margin to a four point margin...
...McCain gains just Missouri and North Carolina.
...Obama's list of states holds steady.
...the total number of toss up states drops from 12 to 10.

If the line between lean and toss up is downgraded to the 3 point mark...
...McCain shifts Florida and Montana into his lean category.
...Obama adds Michigan as a lean state.
...the total number of toss up states decreases from 12 to seven.

The likelihood of the five states in question switching sides is diminishing, but I don't know whether the window is completely shut either. The question, then, is is this a good time/the best time to make this move with the debates coming up? It is a nice line of demarcation in the events on the ground, but it may not offer the best time for such a transition. Thoughts?

I should add that I'm inclined to go ahead and move the line down to three and leave it there for the duration of the race. If any of these states are going to be toss ups they'll likely move in that direction anyway.


Recent Posts:
Today's Agenda

The Electoral College Map (9/21/08)

The Electoral College Map (9/20/08)

9 comments:

Robert said...

I'll junp in. I would suggest that you lower it to 4 points now and then to 3 points after the last debate. I suspect there will be some movement resulting from the debates.

Unknown said...

I'm a little hesitant to close it too far. When we say that states within 5 points are toss-ups we are partly acknowledging that there could be a 5 point swing between now and the election, and that probability gets smaller as time goes on. But we're also saying that the polls may be getting it systematically wrong, because, for example, of GOTV efforts.

Most of the ground game estimates I've seen, though, put the realistic limit of turnout efforts at about a 2% shift in the results. So Robert's suggestion of going to 4% after the first debate and 3% after the last debate seems good to me.

Will you also change the lean/solid line while you're at it? It seems even less likely that a 9% state is even remotely in play now. How about setting the lean/solid line at double the toss-up lean line?

Unknown said...

What's the margin of error on most of these polls? That's what I'd set the threshold between "lean" and "toss-up" as. If it's between percentage points, I'd round up (i.e., if the typical margin of error is anywhere from 4.01% to 4.99%, I'd use 5% as the cutoff).

Anonymous said...

A good set of comments here.

Scott,
I agree with you on scaling the category back too far. But I'm also resistant to constantly changing things. As you have pointed out, it is the continuity here that helps make this site distinctive. That said, two thoughtful and reasoned shifts down the stretch is probably not too radical a shift.

So Rob, I think you're on to something with this approach you've suggested. I think what I'll do is bump the mark down to four on Friday night and then again to three on Friday night October 17 at the close of the news week following the final debate.

On the lean/solid divide:
I'm going to have to think about this one some. The reality is that this race is going to shift away from these degrees of difference to a competitive/non-competitive differentiation. However, I'd like to keep that middle category just for the sake of enhancing our understanding of the natural order of states in the race. But I think that will require a bit of maintenance. I want to maintain the integrity of the lean category, but I don't want it to appear as if I'm moving the line just for the sake of keeping it in the mix. I may leave the line where it is and see what direction things move in before making a decision there. But I do like this double the value of the toss up/lean line idea. I'll crunch the numbers a few different ways ahead of Saturday's update and see what seems most appealing and/or makes the most sense. It doesn't look like the distinction makes sense when you get down to 19 electoral votes in one category, so I'll want to try and stay ahead of that as best I can.

S.D.,
The margins in these state polls range anywhere from around 3-5 points. That's why three is the lowest I will likely go when setting that particular line.

Jack said...

I'm going to second what Robert, Scott and Josh have said on the tossup/lean boundary. On the lean/strong, my completely unscientific suggestion is to set it at 8 now or after the debates and keep it there.

Anonymous said...

Unscientific? There's no room around here for that!

Eight seems like a logical line, though. I'll have to post some of the different outcomes this weekend after the switch.

Anonymous said...

Oh Scott, I'll have your map up tomorrow. If you have any additional analysis (outside what you added to the original comment), feel free to email me by clicking "contact" under my profile information.

Robert said...

Josh,

Since you are averaging several polls, doesn't the margin of error decrease? Obviously, the averages from those states polled many times would be more reliable than averages from states polled only a few times.

Anonymous said...

That's right, Rob. There is some amount of canceling out of the margin of error going on with an average of all these polls. But as it is weighted, obviously there is some extra weighting on that most recent poll's margin of error. Still, if this is the limbo, 3 is as low as I'll go.