Tuesday, August 19, 2008

Is Rasmussen's Inclusion of "Leaners" Affecting the Electoral College Outlook Now? An Update

It has been about three weeks since FHQ last checked in on the effect the early July decision by Rasmussen to report "leaners" in their polling releases had on our electoral college projections. At the time, only Ohio switched from favoring Obama to favoring McCain based on which version (with leaners or without) was used. In the period since though, there have been 27 new Rasmussen polls in 27 states. For 18 of those, that means the second (or third in the case of Minnesota) poll since the switch. The point of the comparison between the two types of numbers initially was to account for the discrepancy in the comparison of with leaners polls to previous without leaners polls. It just wasn't an apples to apples comparison.

With multiple "leaners" polls now out in 18 states, that really isn't the concern anymore. There are still nine states that had their first "with leaners" results reported since the beginning of August (and 13 others that have yet to have had their first with leaners poll conducted), but this is now primarily an exercise in examining how the electoral college projection would change -- and not a critique of the wholesale switch over to using the "leaners" numbers that took place in the aftermath of the change. We have the data, why not look at it? The more information we have from a "with leaners"/"without leaners" perspective, the more likely we are to begin seeing differences in the electoral college projections.

What we saw initially was that the leaners were breaking for McCain nearly across the board, but that around the time of Obama's trip abroad, that began to switch. And since that point, McCain's advantage in the "with leaners" numbers has almost been cut in half (from +1.08 to just +0.64 now). [Since the table is getting so large, I've decided to just simply append it to the end of the post instead of breaking up the text with such a mammoth figure. The new data are below the re-labelling of the columns.] Regardless, McCain continues to maintain an advantage over Obama with the leaners in these polls. And in an election that looks like it could come down to who is swinging the most independents, that's an important distinction.

But let's look at that more closely, focusing on the more recent polls (the new additions this time). Of those 27 states, eight are red states, eleven are blue states and the remaining eight are toss ups. In nine of the eleven blue states, the leaners are going for Obama (Though, it should be noted that in five of those nine, the leaners margin is exactly the same as the without leaners margin.), and in five of the red states, the leaners move in McCain's direction. So McCain in red states and Obama in blue. No real surprise there.

In the toss up states though, there are some differences. McCain holds slight advantages in 5 of those eight toss up states. The leaners broke for him in 60% of those cases (Missouri and Montana being the exceptions.). The picture for Obama was different. Of the three toss ups that favored the Illinois senator, the leaners moved toward him in one (Colorado), against him in another (Michigan) and broke even in the third (Nevada). That list of states was the very same as the one I discussed yesterday in the post about the benefits McCain could gain from tapping Mitt Romney as his running mate. Those three plus all the states in various shades of red add up to 271 electoral votes. And it is a very small consolation that Obama gains leaners in Colorado yet has the overall lead in the poll go to McCain. Nor is it beneficial to break even in Nevada when the overall poll favors McCain -- a switch from the month prior. In other words, having leaners break for him in Missouri and Montana is about all Obama can hang his hat on in this instance.

But what about the electoral college projection? Well, no matter which measure you use -- with or without leaners -- the projection (as a function of our weighted average) remains the same: 298-240 in favor of Obama. In fact, it is interesting that some of these leaner gains are cancelling each other out when subsequent polls are released. That has happened in both Minnesota and Wisconsin. Also, Nevada and Oregon have shown the same margins in both versions of the polls across two releases now.

One additional pattern that we can glean from these Rasmussen polls is the line up of states that are likely to have new polling data released in the next week. If everything holds to form, there should be new data in Arkansas, California, Florida, New Hampshire, New Mexico, Ohio and Pennsylvania in the near future. The last five of those are certainly worth keeping an eye on. All are toss ups with the exception of New Mexico. Given how other western states favoring Obama (Colorado and Nevada) have fared lately, it will be interesting to see if New Mexico follows suit.

Rasmussen Polls Since w/Leaners Distinction was Added (7/9/08)*
StateDatew/o Leaners
w/Leaners
Change
Undecideds Drop
Missouri
7/7+5
+50
-10
New Jersey
7/7+5
+3+2
-4
Illinois
7/8+13
+11+2
-6**
North Dakota
7/80
+1+1
-7**
Wisconsin
7/8+13
+10+3-6
Louisiana
7/9+20
+19+1
-2
South Dakota
7/9+4
+40
-4
Washington7/9+9
+8+1-6
Iowa
7/10+10
+100
-9
Michigan
7/10+8
+80
-5
Minnesota
7/10+18
+17+1
0
Kansas
7/14+20
+23+3
-9
North Carolina
7/15+3
+30
-5
Oregon
7/15+9
+90
-5
Nevada
7/16+2
+20
-5
Virginia
7/160
+1
+1
-6
Alaska7/17+5+5
0
-7
Arkansas7/17+10
+13+3-4
Georgia7/17+9
+11+2
+1
Maine
7/17+10
+8+2
-2
Colorado
7/21+7
+3+4-13
Ohio
7/21+6
+10+4
-7
Florida
7/22+1
+2+1-7
Minnesota
7/22+12
+13+1--***
New Hampshire
7/23+6
+4+2-5
Pennsylvania
7/23+5
+6+1-8
New Mexico
7/24+5
+6+1-6
California
7/24+12
+10+2-6
StateDatew/o Leaners
w/Leaners
Change
Undecideds Drop
Mississippi
7/28+11
+12+1-2
Nebraska
7/28+18
+19+1-3
Kentucky
7/29+10
+9+1-9
Montana
7/29+1
0+1-3
Alaska
7/30+5
+6+1--***
Arizona
7/30+16
+19+3-6
Texas
7/30+9
+8+1-7
Alabama
7/31+18
+20+2-9
Connecticut
7/31+15
+13+2-6
New Jersey
8/4+8
+10+2--***
New York
8/4+20
+19+1-2
Massachusetts
8/5+15
+16+1-6
Wisconsin
8/5+4
+7+3--***
Washington
8/6+12
+120
--***
Iowa
8/7+5
+50--***
Michigan
8/7+7
+4+3--***
Missouri
8/7+7
+6+1--***
Oregon
8/7+10
+100
--***
Kansas
8/11+15
+14+1--***
Nevada
8/11
+3
+30
--***
Illinois
8/12+15
+150
--***
Maine
8/12+13
+14+1--***
Virginia
8/12+1
+1+2--***
Colorado
8/13
+2
+1+1--***
Minnesota
8/13
+4
+40
--***
North Carolina
8/13+4
+6+2--***
Georgia
8/14+7
+9+2--***
Avg. Change+0.64-5.57
*The "with leaners" distinction was added to reports that were released beginning on 7/9/08. The date on which these polls were conducted (The ones that these releases were based on) stretches back to 7/7/08.
**Rasmussen has only conducted one poll in these states. Therefore, the difference was taken from between the with and without leaner numbers within the same poll in these cases.

***Previous poll had been taken after "with leaners" change had been made.



Recent Posts:
The New Ohio Poll and McCain's VP Choice

The Electoral College Map (8/17/08)

Which States are Underpolled in the Presidential Race?

Monday, August 18, 2008

The New Ohio Poll and McCain's VP Choice

With Public Policy Polling's release of a new poll in Ohio last night, the current state of the presidential race took on a much closer feel. Let's look at yesterday's Electoral College Spectrum with that poll incorporated to get a better feel for the dynamics as they now stand.

The Electoral College Spectrum*
HI-4
(7)**
WA-11
(165)
PA-21
(264/295)
FL-27
(369/196)
LA-9
(67)
VT-3
(10)
MN-10
(175)
CO-9***
(273/274)
NC-15
(384/169)
ID-4
(58)
RI-4
(14)
DE-3
(178)
OH-20
(293/265)
SC-8
(154)
NE-5
(54)
MD-10
(24)
OR-7
(185)
NV-5
(298/245)
SD-3
(146)
WY-3
(49)
IL-21
(45)
NJ-15
(200)
VA-13
(311/240)
TX-34
(143)
AR-6
(46)
CT-7
(52)
IA-7
(207)
ND-3
(314/227)
GA-15
(109)
TN-11
(40)
NY-31
(83)
WI-10
(217)
IN-11
(325/224)
MS-6
(94)
KY-8
(29)
ME-4
(87)
NM-5
(222)
MT-3
(328/213)
WV-5
(88)
AL-9
(21)
CA-55
(142)
MI-17
(239/316)
MO-11
(339/210)
AZ-10
(83)
UT-5
(12)
MA-12
(154)
NH-4
(243/299)
AK-3
(342/199)
KS-6
(73)
OK-7
(7)
*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 New Hampshire (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.
***Colorado is the state where Obama crosses (or McCain would cross) the 270 electoral vote threshold to win the presidential election. That state is referred to as the victory line
.

Wait, that looks exactly like yesterday's ECS. It does. That tie in the Buckeye state made the margin between Obama and McCain smaller there, but Ohio still slightly favors Obama. More importantly though, Colorado, Nevada and Ohio have inched closer to McCain of late, and those three states along with Virginia are the states upon which this race appears to be hinging at the moment. If either candidate gets all their current states outside of that four state block, they will need some combination of those four to break 270 electoral votes.

Here's where McCain's choice of a running mate comes into play. The Arizona senator could play offense in Colorado and Nevada by tapping Mitt Romney as his vice presidential nominee. The former Massachusetts governor would also help in his home state of Michigan. That potentially puts Obama in a real bind. McCain would be on the offensive in those two western states and Michigan and would have to play defense to hold on to Virginia, but wouldn't even "need" Ohio. Granted, if you look at the Spectrum, ceding Ohio and swinging Michigan to the right would cost McCain three electoral votes, but it would help him squeak by in the electoral college with a 271-267 advantage. And hey, Ohio is trending the Arizona senator's way.

Looking at the electoral college breakdown and vice presidential selection in that light makes me second-guess the pundits' thoughts about McCain's pro-choice running mate trial balloon in the Weekly Standard last week. Did that comment indicate Ridge and/or Lieberman or was it referring to someone who has moderated his views on the abortion question, Mitt Romney? There are certainly other issues surrounding Romney, but would his religious background affect the ticket enough to swing any states Obama's way? My first impression is that it would not, though some states (especially in the South) would be closer than they have been in the past. In the end, that is a razor-thin electoral college margin, but at this point, it looks like Romney may be able to do the most damage to Obama and the Democrats in the electoral college. What he brings to the table, is what helps McCain the most in the electoral college.

This swoon period will end for Obama with the effect the combination of a VP selection and the convention helping. Will that effect be muted by the GOP convention that follows on the heels of the Democrats', though? Suddenly, August looks to be a challenging month for the second consecutive cycle for the Democrats. If the race looks tied after whatever bounce the GOP convention gives McCain, then the debates will likely play a crucial role in deciding who will win this election in November.

Speaking of convention bumps, Thomas Holbrook has a post up on his blog now that looks at the effects of past conventions and glances ahead to the upcoming conventions.


Recent Posts:
The Electoral College Map (8/17/08)

Which States are Underpolled in the Presidential Race?

The Electoral College Map (8/14/08)

Sunday, August 17, 2008

The Electoral College Map (8/17/08)

The latter half of the week closed with just a handful of new polls. That is quite a departure from recent weeks when the bulk of polling releases had been backloaded, occurring between Wednesday and Friday (No, there have not been many polls released over the weekend since primary season ended, and yes, I did put Wednesday's polls in the mid-week update due to the delay caused by my foray into the 2004 data.). Still, even with just six new polls from five states out, there were some shake ups throughout FHQ's various depictions of the race's dynamics.

New Polls (Aug. 14-17)
StatePollMargin
(With Leaners/ Without Leaners)
Colorado
Rocky Mountain News
+3
Colorado
Rasmussen
+1/+2
Maine
Rasmussen
+14/+13
Minnesota
Rasmussen
+4/+4
North Carolina
Rasmussen
+6/+4
Texas
University of Texas
+10

The first impression is that that is a lot of red. Must be good for McCain, right? Yes, but maybe not for that reason. North Carolina and Texas are a toss up and a lean state, respectively and both continue to hold steady in those positions favoring McCain. The same is true of Maine, though the Pine Tree state is blue, not red. With the new polls in Colorado and Minnesota, though, we see a bit of a divergence from what we have become accustomed to in both. Colorado is beginning to look a lot like Nevada: a western state that remains blue but is trending in McCain's direction. Each have been and continue to be toss up states and the polls in each reflect that. An overwhelming majority of the Colorado polls released since Obama clinched the Democratic nomination have been within the margin of error, but since mid-July half of the six polls have favored McCain. That's departure from the pattern that had dominated before that: close polls favoring Obama. And Minnesota? The North Star state also saw a change from typical polling patten.

Changes (Aug. 14-17)
StateBeforeAfter
Minnesota
Strong ObamaObama lean

Minnesota becomes the first blue state to move away from Obama into another category since Ohio turned pink based on a Rasmussen poll in late July. But Ohio is a toss up state. We'd expect, to some extent, a toss up state to be more volatile than a state that is either a lean or strong. Minnesota was actually the last such state to move away from Obama all the way back on June 3. [Yeah, the day Obama wrapped up the nomination.] I glanced back through the maps to the point where I adopted the weighted average on April 30, and Minnesota -- a state that has hovered around the line between strong and lead all along -- was the only blue state (lean or strong) to have moved away from Obama in that time. By comparison, McCain has had 6 lean or strong states (Alaska, Florida, Montana, North Carolina, South Dakota and Texas) move away from him since mid-June (not counting the states that shifted in one direction and moved back). Just two of those six (Alaska and South Dakota) have come since mid-July, though.
[Click Map to Enlarge]

So, while some states have shown signs of trending toward McCain of late, that movement has yet to manifest itself in the electoral vote tally. The electoral college still favors Obama by a 298-240 margin. Obama, though, is now 10 electoral votes down in his strong category. The total of Obama strong state electoral votes remains larger than the sum of electoral votes in both McCain's strong states and the states leaning in his direction. That cushion is not as big anymore, though. And while that isn't readily apparent on the map above, we can begin to see it in the Electoral College Spectrum (ECS).

The Electoral College Spectrum*
HI-4
(7)**
WA-11
(165)
PA-21
(264/295)
FL-27
(369/196)
LA-9
(67)
VT-3
(10)
MN-10
(175)
CO-9***
(273/274)
NC-15
(384/169)
ID-4
(58)
RI-4
(14)
DE-3
(178)
OH-20
(293/265)
SC-8
(154)
NE-5
(54)
MD-10
(24)
OR-7
(185)
NV-5
(298/245)
SD-3
(146)
WY-3
(49)
IL-21
(45)
NJ-15
(200)
VA-13
(311/240)
TX-34
(143)
AR-6
(46)
CT-7
(52)
IA-7
(207)
ND-3
(314/227)
GA-15
(109)
TN-11
(40)
NY-31
(83)
WI-10
(217)
IN-11
(325/224)
MS-6
(94)
KY-8
(29)
ME-4
(87)
NM-5
(222)
MT-3
(328/213)
WV-5
(88)
AL-9
(21)
CA-55
(142)
MI-17
(239/316)
MO-11
(339/210)
AZ-10
(83)
UT-5
(12)
MA-12
(154)
NH-4
(243/299)
AK-3
(342/199)
KS-6
(73)
OK-7
(7)
*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 New Hampshire (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.
***Colorado is the state where Obama crosses (or McCain would cross) the 270 electoral vote threshold to win the presidential election. That state is referred to as the victory line
.

Minnesota's shift doesn't look like all that big a deal now. In fact, it hasn't moved at all. The North Star state changed colors but maintained the same position in the ECS. What is different is that, for the first time, a state other than Pennsylvania is in the Victory Line slot. That more accurately reflects what is (and has been) happening in the race. Pennsylvania has been trending toward Obama while Colorado has not. While not necessarily favoring McCain, the margins in the Centennial state have drawn closer to zero. With Colorado and Pennsylvania basically switching places, the result is that Obama's path to 270 is not as clear. If Colorado and Nevada are trending toward McCain (They are both still in blue above the Partisan Line.) that makes Ohio that much more important. If both western states turn pink, Ohio is the state that would put Obama over the top. Without those two western states and Ohio, McCain becomes president. In fact let's look at it this way: if Colorado and Nevada move into McCain's column and Ohio holds its position, the Victory and Partisan line would converge on the Buckeye state's spot in the ECS. In other words, if the election played out that way, we would basically have a replay of the 2004 election. And this election was supposed to be so much different than those before it. It may yet be, but with the way things are shaping up at the moment, we're looking at another close election with the map changing very little.

The Watch List*
StateSwitch
Alaska
from Toss Up McCain
to McCain lean
Floridafrom Toss Up McCain
to McCain lean
Georgiafrom McCain leanto Strong McCain
Minnesotafrom Obama lean
to Strong Obama
Mississippifrom Strong McCainto McCain lean
Nevadafrom Toss Up Obamato Toss Up McCain
New Mexicofrom Obama leanto Toss Up Obama
North Carolinafrom Toss Up McCain
to McCain lean
Ohiofrom Toss Up Obamato Toss Up McCain
Virginiafrom Toss Up McCainto Toss Up Obama
Washingtonfrom Strong Obamato Obama lean
Wisconsinfrom Obama leanto Toss Up Obama
*Weighted Average within a fraction of a point of changing categories.

So what should we be watching for in the next week? New polling in any of the twelve states above could potentially bring about the changes charted in the Watch List. Minnesota is the only change on the list since Thursday. But that tighter Rasmussen poll didn't shift the state enough to pull it firmly into the lean category. The North Star state continues to oscillate relatively tightly around the line between the lean and strong categories. Finally, even though Colorado jumped both New Hampshire and Pennsylvania in the ECS, the average still isn't close enough to warrant its inclusion on the Watch List. If the Centennial state keeps trending toward McCain, though, it will work its way to that point.


Recent Posts:
Which States are Underpolled in the Presidential Race?

The Electoral College Map (8/14/08)

2008 vs. 2004, Part II: What Happened in the Final 100 Days in 2004 and What That May Mean for the Rest of This Campaign

Friday, August 15, 2008

Which States are Underpolled in the Presidential Race?

I mentioned in yesterday's electoral college post that I thought Nevada was a state that had been polled less often than than it should have been given how closely contested the Silver state appears to be. Of course I was called on to elaborate on that assessment.* Jack may have been asking for a simple gut reaction as to what I considered to be underpolled. However, I cannot help but over-analyze even the simplest of questions. Why provide a feeling when we can put the data we have to good use?

With that in mind, what is underpolled?

We can go about answering that question in a couple of ways. The simplest way is to take an average. With the release of Rasmussen's poll in North Carolina this morning, the grand total of polls in our data set (from Super Tuesday to now) is 553 polls. That's an average of just over 11 polls per state. States coming in under that line, then, are underpolled. Sure, that certainly isn't false, but that is rather a low bar to set in defining what "underpolled" means.

Another couple of layers can also be added to this. We would expect that the number of polls conducted in a state would vary based upon how close and how large the state was. We'll get to a state's size in a moment, but let's focus initially on the "how close" question. An easy way to extend the simple approach is to split the states into groups according to how close they are. Well, that's already been done for us. We can take an average of the toss up states, the lean states and the strong states with the expectation that toss up states would have more polls conducting in them on average than a lean state or a strong state (Likewise, lean states would have more polling than strong states.).

Average Number of Polls in States by Level of Competitiveness

StatesPoll Frequency
Average
Toss Ups
(14)
AK CO FL IN MI MO MT NV NH NC ND OH PA VA
218
15.53
Leans
(10)
DE GA IA NJ NM OR SC SD TX WI
116
11.6
Strongs
(26)
AL AZ AR CA CT HI ID IL KY KS LA ME MD MA MN MS NE NY OK RI TN UT VT WA WV WY
219
8.42

And that is what we see in the table. So instead of saying the overall average of polls across all 50 states is 11 and there have only been 10 polls in Nevada. We can say that among toss up states, the average number of polls is 15.5 and Nevada has had only 10 polls conducted since February. That gives us a better definition of underpolled.

It gives us a better definition, but perhaps not a very efficient one. What about state size? We'll get to that in a minute. First, we can take a page out of FiveThirtyEight's book and run a regression with the number of polls conducted so far in a state as the dependent variable and the competitiveness that state (as measured by our weighted average) as our explanatory variable. In other words, we would expect that as the spread between the two candidates increases, the number of polls in that state decreases. That's exactly what the graph below depicts.

Predicted Polling Frequency
[Click Graph to Enlarge]

And with that handy regression line, we can predict where a state's frequency of polling should be given its level of competitiveness. So, Nevada, with ten polls thus far is about six polls under what we would expect in light of how close the race appears in the Silver state. But right there in that lower left quadrant of the graph are several toss up states clustered together. Alaska, Indiana, Montana, New Hampshire and North Dakota all come in under that prediction line. Even lean states like New Mexico, South Carolina and South Dakota are underpolled.

And what about a state's size? The number of electoral votes at stake in a state -- a reasonable proxy for size in this context -- could affect the frequency of polling in a state as well. When we add that into the regression how are the things we see above affected? Again, that would add to our understanding of what is causing polling frequency to vary across states and ultimately increases the efficiency of our prediction. Competitiveness alone explains about a quarter of the variation in polling frequency and competitiveness and state size bumps that up to just over half. If we focus our attention on the 14 toss up states -- six of which were underpolled when compared to the original prediction -- only four were significantly underpolled: Indiana, Montana, Nevada and North Dakota. Alaska, Michigan and New Hampshire were about on par with where they would be predicted to be with 10, 17 and 13 polls, respectively. The remaining seven states could be considered "overpolled" based on competitiveness and state size. You cannot over poll in my opinion, but in a world of finite resources and comparatively speaking, that's the reality.

So, long story short, it is that small group of toss up (and some) lean states that are underpolled at the moment.


*Our readers and commenters here are great. I certainly have my own ideas of what to post here, but it is in my conversations both here in the comments section and with colleagues here at UGA that spur some of the great ideas that ultimately appear in this space. I don't say it often enough, but thank you all for your support of the site and for your contributions.


Recent Posts:
The Electoral College Map (8/14/08)

2008 vs. 2004, Part II: What Happened in the Final 100 Days in 2004 and What That May Mean for the Rest of This Campaign

2008 vs. 2004, Part I: What Things Would Have Looked Like 4 Years Ago This Time

Thursday, August 14, 2008

The Electoral College Map (8/14/08)

The upside to waiting until late in the day to update our electoral college projections is that there is a chance to incorporate all or most of the polls from that day. The downside is that you have to incorporate all those new polls. And when you are already behind adding in the new polls into the existent averages, you make meeting a goal of "I'll have an update up later tonight" that much more difficult to attain. [At least I had that 2004 material up. 2004!?!] Regardless of the timing, there have been 17 new polls released in 14 states since Sunday. Six of those 14 states are toss ups by FHQ's measure and one more state moved into that category with the addition of new polling.

New Polls (Aug. 10-13)
StatePollMargin
(With Leaners/ Without Leaners)
Alaska
Moore
+3
Alaska
Hays
+5
Colorado
Public Policy Polling
+4
Florida
Insider Advantage
+4
Iowa
Rasmussen
+5/+5
Kansas
Rasmussen
+14/+15
Kentucky
Survey USA
+18
Nevada
Rasmussen
+3/+3
New Jersey
Quinnipiac
+10
North Carolina
Survey USA
+4
Oregon
Rasmussen
+10/+10
Pennsylvania
Franklin & Marshall
+8 /+5
(reg.)/(likely)
Virginia
Insider Advantage
0
Virginia
Rasmussen
+1/+1
Virginia
Survey USA
+1
Washington
Survey USA
+7
Wisconsin
Strategic Vision
+5

That should serve as a hint that, surprisingly, it was not Nevada that moved over to McCain as a result of the Rasmussen poll that gave the Arizona senator a slight lead in the Silver state. And hey, it was less than 24 hours ago that I lamented the fact that Nevada had not been polled in almost a month. I still think that Nevada has been under-polled given how close it appears. Let's sink some of the resources that have been sunk into (over-)surveying New York and put that into Nevada. The Empire state ain't movin' folks. I expect it to be an easy win for Obama on November 4. Nevada, however, is a close state. By our measure, it is now the closest state, surpassing Ohio to claim that distinction. But I'll save that discussion for when we get to the Electoral College Spectrum below.

Changes (Aug. 10-13)
StateBeforeAfter
Alaska
McCain leanToss Up McCain

With that small rant over, let's shift the focus to Alaska. The Last Frontier (I'm not a Star Trek fan, but I have to fight my subconscious every time I type that. Alaska is not, in fact, the Final Frontier.) on the weight of two new polls moves into the toss up category. That Ivan Moore poll from late July slipped under the radar somehow, but at least I wasn't the only one to miss it. Of course Nate Silver is reporting that poll as having a 2.4 point margin. His link to the poll is dead now and everyone else seems to be reporting it as a 3 point margin. Pushing that point aside for the moment, Alaska is now close enough to be considered a toss up. Yikes if you're McCain. This is a state that Bush won by 25 points in 2004. That's over 20 points that Alaska as slid since November 2004. Now, just to be clear, Alaska still favors McCain but has inched closer to Obama based on these two polls. Bush's approval (of lack thereof) and the Stevens' indictment have put a distinct blemish on the Republican brand in the state and that has translated into the presidential trial heats drawing closer together. [Just for the sake of transparency -- or at least accurate reporting -- I should note that the Moore poll was released a full week before that indictment was handed down.]
[Click Map to Enlarge]

Alaska moved and Nevada didn't, so the shift I expected when I saw the new Silver state poll (but before I imputed it into the state's weighted average) didn't actually happen. As a result the electoral vote tallies for each candidate remain unchanged. McCain, though, loses three more electoral votes from the lean plus strong state total that is, by our estimation, safer for the moment. That total comes to 154 electoral votes for McCain which is over twenty electoral votes fewer now than what Obama has in just his strong category (175 EVs). And that brings into even starker contrast the difference between where McCain-Obama stand in regard to the Electoral College Spectrum (ECS) and where Bush and Kerry were four years ago at this time. The intensity of support for Bush has certainly given way to a far different picture for the Republican standard bearer in 2008. If Obama can force McCain to play defense in those eight states (86 EVs), McCain will be at a distinct disadvantage as we near the heart of this campaign. I have no idea what Obama's schedule is coming off his vacation -- he does have a running mate to name -- but would a quick trip to Alaska be that out of the way? That sort of move could turn 5 point poll leads in the state into more than just an outlier. Granted, driving back from my own vacation in late July, I saw a truck with Alaska tags that sported a "NObama!" bumper sticker. That isn't representative of the entire state but I'm willing to err on the side of caution and give the GOP some credit in a state that has only voted for a Democratic presidential candidate once in its history (LBJ in 1964).

The Electoral College Spectrum*
HI-4
(7)**
WA-11
(165)
NH-4
(252/290)
NC-15
(357/196)
LA-9
(67)
VT-3
(10)
MN-10
(175)
PA-21***
(273/286)
FL-27
(384/181)
ID-4
(58)
RI-4
(14)
DE-3
(178)
OH-20
(293/265)
SC-8
(154)
NE-5
(54)
MD-10
(24)
OR-7
(185)
NV-5
(298/245)
SD-3
(146)
WY-3
(49)
IL-21
(45)
NJ-15
(200)
VA-13
(311/240)
TX-34
(143)
AR-6
(46)
CT-7
(52)
IA-7
(207)
ND-3
(314/227)
GA-15
(109)
TN-11
(40)
NY-31
(83)
WI-10
(217)
IN-11
(325/224)
MS-6
(94)
KY-8
(29)
CA-55
(138)
NM-5
(222)
MT-3
(328/213)
WV-5
(88)
AL-9
(21)
ME-4
(142)
MI-17
(239/316)
MO-11
(339/210)
AZ-10
(83)
UT-5
(12)
MA-12
(154)
CO-9
(248/299)
AK-3
(342/199)
KS-6
(73)
OK-7
(7)
*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 Colorado (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.
***Pennsylvania is the state where Obama crosses (or McCain would cross) the 270 electoral vote threshold to win the presidential election. That state is referred to as the victory line
.

Alaska jumps both Florida and North Carolina on the ECS in the process of becoming a toss up, but with so many new polls being released there were a few shake ups, though, nothing major. Nevada is so close to being a tie now that it jumped Ohio, becoming the least blue state. Those two, along with Virginia are the most closely contested states at this point according to our weighted averages. And even with three polls, Virginia held steady; now firmly within the area that would put the state on the Watch List. Outside of that Pennsylvania is the only other state of note. The Keystone state nearly ceded its position as the Victory Line to New Hampshire. The two are separated by only two one-hundredths of a point. However, I should add one caveat since Pennsylvania is part of the discussion. Franklin and Marshall reported both their likely voter numbers and registered voter numbers in their Pennsylvania poll. Most are reporting that poll as a 5 point Obama advantage, but FiveThirtyEight has decided to use the registered voter numbers and is reporting the 8 point margin among that sample group. Here's the thing: I don't particularly want to wade into this debate here, but I will be open about how each of those results affects the average. With the likely voters 5 point margin, Pennsylvania is not as close to New Hampshire's average (two tenths of a point instead of two one-hundredths), but the 8 point, registered voter margin pushes the commonwealth's average right up against the Granite state's.

The Watch List*
StateSwitch
Alaska
from Toss Up McCain
to McCain lean
Floridafrom Toss Up McCain
to McCain lean
Georgiafrom McCain leanto Strong McCain
Minnesotafrom Strong Obamato Obama lean
Mississippifrom Strong McCainto McCain lean
Nevadafrom Toss Up Obamato Toss Up McCain
New Mexicofrom Obama leanto Toss Up Obama
North Carolinafrom Toss Up McCain
to McCain lean
Ohiofrom Toss Up Obamato Toss Up McCain
Virginiafrom Toss Up McCainto Toss Up Obama
Washingtonfrom Strong Obamato Obama lean
Wisconsinfrom Obama leanto Toss Up Obama
*Weighted Average within a fraction of a point of changing categories.

Pennsylvania, though, is comfortably within the range of toss up states, but it is trending away from the Partisan line and thus not close enough to be included on the Watch List. The states most likely shift to categories (or across partisan distinctions) have added two states since Sunday. Alaska's new polls put the state on the Watch, but only barely so. Actually, Alaska's presence on the list depends on whether you use the 3 point margin in the Moore poll or the 2.4 point margin (mentioned above) that Nate Silver has reported. The 2.4 point margin would lower the average enough to pull Alaska off the list entirely. As it is, with the three point margin, the Last Frontier is right on the line between categories. Wisconsin slips on to the Watch as well. That five point Obama advantage in the latest Strategic Vision poll out of the Badger state continues a pattern of tighter, yet comfortably Obama, polls. Wisconsin has hovered around the line between lean and toss up for most of this cycle, so this isn't a shock. The state is getting tighter though.

At some point I'd like to address how wide the gaps are between some of these categories, but that point isn't today. Since I've brought in the Wednesday polls to this update, I may have created less work for myself on the weekend update. If that proves to be true I'll make some statements about those spreads then. It would certainly help our understanding of both the Electoral College Spectrum and the Watch List.


Recent Posts:
2008 vs. 2004, Part II: What Happened in the Final 100 Days in 2004 and What That May Mean for the Rest of This Campaign

2008 vs. 2004, Part I: What Things Would Have Looked Like 4 Years Ago This Time

The Electoral College Spectrum

Wednesday, August 13, 2008

2008 vs. 2004, Part II: What Happened in the Final 100 Days in 2004 and What That May Mean for the Rest of This Campaign

Now that we've seen that John Kerry's state-level poll leads over Bush in 2004 were similar in the aggregate to the electoral college vote distribution we see currently in the 2008 race, we can focus more on what appears to have happened over the final 100 or so days of the race four years ago. By extension, then, we may be able to make some educated guesses as to what we may have in store between now an November. Here again is the map that shows how the electoral college would have looked in mid-August if FHQ's weighted average methodology were used on the polling from 2004.
[Click Map to Enlarge]

Based on Florida (or at least on Kerry's edge in the average there in August), Kerry held an advantage over the incumbent president in the August electoral college projection. Florida, though, wasn't the only mover between August and November. It was however, among a group of states that shifted their electoral votes from one side of the partisan line to the other -- Iowa and New Mexico were the other two. Below are the final results from the 2004 election. Each state was color coded based on the final vote percentage margin four years ago. California, for example, was firmly within Kerry's column throughout 2004, but the final percentage margin (9 points) put the Golden state in a position to be deemed a Kerry lean state. While Florida, Iowa and New Mexico were the only three states to have shifted sides, there were twelve overall movers when comparing the August projection to the final results.
[Click Map to Enlarge]

Let's isolate those twelve states just to get a visual on where exactly the movement occurred. Not only do Florida, Iowa and New Mexico turn red, but several states became more intensely red and all twelve moved toward Bush including the two blue states on the map, California and Minnesota. The momentum seems to have been squarely on the incumbent president's side coming down the stretch. Bush made his biggest gains -- moving across two categories -- in Florida and New Mexico and seems to have outperformed state-level polling average through August most consistently in the peripheral South and in border states.
[Click Map to Enlarge]

That momentum translated into a shift of 39 electoral votes over the final three months of the 2004 campaign and into an electoral college victory for Bush. So what does any of this have to do with the current race for the White House? It should serve as a cautionary tale for all you President Kerry enthusiasts and Obama supporters. Much can change and likely will over the course of the next two plus months, and though, in 2004 that meant a shift toward the incumbent, in 2008 it may mean something different. One pattern that has emerged in this current race is that things often return to normal after a "shock" to the system. Obama wins the Democratic nomination and gets a bounce that trailed off in some states. McCain punches back while Obama is abroad and upon his return and draws closer to the Illinois senator. But even that has given way to a regression toward the mean. You see different pictures of this depending on where you look. Real Clear Politics shows Obama rising ever so slightly while McCain dips since last week. Over at Pollster the picture is a bit different with both candidates trending upward, but McCain is doing so at a quicker rate.

Granted that focuses on the national picture and Alan Abramowitz has already covered that. One thing that we can note of 2004 is that half of those movers from August on were toss up states. We can then feel comfortable that that is where most of the action will be from now on. That's 13 states as of FHQ's most recent accounting of the electoral college. [I'll have an update up tonight. Yeah, I know -- late.] Of the big hitters, Pennsylvania is trending toward Obama, Virginia and Ohio are holding steady but very close, Michigan and Colorado are closing a bit, and Florida, after moving hard toward Obama has held steady like North Carolina at a distance just outside of Obama's reach. One thing that should be noted is that both Indiana and Nevada have not been polled very recently and are certainly in need of updating. With that said, one thing that struck me in going over the data from 2004 was how big the uptick in polling was once September started. We have been lucky to have had as much information during this cycle as we've had, but it should really start to pick up after the GOP convention during the first week in September.

I should also put in a request to all FHQ readers. I have access to an incomplete set of the polling from 2000 and would like very much to do a similar sort of analysis with that data. If anyone knows where that data can be found just let me know in the comments section below. Thanks.


Recent Posts:
2008 vs. 2004, Part I: What Things Would Have Looked Like 4 Years Ago This Time

The Electoral College Spectrum

The Electoral College Map (8/10/08)

2008 vs. 2004, Part I: What Things Would Have Looked Like 4 Years Ago This Time

Just last week, Alan Abramowitz posted on Sabato's Crystal Ball an examination of how the 2004 presidential race looked in early August. I've been sitting on the state polling data from 2004 for most of the summer, but have yet incorporate it in this forum. I don't particularly like to jump on a bandwagon, but if I've got to follow someone, Abramowitz is a good person to follow up on. His focus was primarily on polling from the national perspective, so an investigation from the state level dovetails nicely with that and augments our understanding of the overlapping dynamics in both races as well as their differences.

In this first part, I want to treat 2004 as if it was the election FHQ was currently examining twice every week. To put it slightly differently, this post applies both the format and methodology of the electoral college projection posts to the state level polling of 2004. Given polling conducted prior to August 12, 2004, what did things look like on August 13 four years ago? Which states were in play? Who held an advantage in the electoral college?
[Click Map to Enlarge]

In the aggregate at least, the Bush-Kerry race looks an awful lot like McCain-Obama does now. Four years ago, Kerry would have held a 44 electoral vote lead over George W. Bush. Over the course of the summer, Barack Obama has had a 298-240 electoral vote edge over John McCain. Notably, Florida and Ohio have switched places in the interim: Florida turning pink and Ohio going light blue. Like Ohio since 2004, Colorado and Nevada have gone blue as well. Other than those four states -- all of which remain toss up states in 2008 -- everything is exactly as it was four years ago.

So should John McCain be feeling pretty good about his position in the 2008 race?

Well, yes and no. It is true that the numbers look eerily similar [Democrats are getting that nervous feeling again.] to the ones in 2004. However, there's more to it than just reds and blues on a map. If we shift our focus to the Electoral College Spectrum, we can get a better idea of exactly how intensely red or blue those states were (ranked from most Democratic to most Republican). This is where the two races look totally different. Bush had at least a 5 point lead (lean or strong states) over Kerry in 22 states with 187 electoral votes. McCain on the other hand isn't as strong overall with strong and lean states (19 of them) totalling 157 electoral votes. In a tight race 30 electoral votes is a big difference. On the flip side, John Kerry's strong and lean state electoral vote tally summed to 198, whereas Obama's total of similar states adds up to 222. From 2004 to 2008, then, the intensity has shifted from the right to the left. But primary season really already told us that. Higher turnout for and a higher number of new registrants participating in the 2008 Democratic primaries are fairly good indicators of that.

The Electoral College Spectrum*
MA-12
(15)**
ME-4
(168)
WI-10
(264/284)
CO-9
(166)
AL-9
(79)
RI-4
(19)
HI-4
(172)
FL-27***
(291/274)
NC-15
(157)
TX-34
(70)
NY-31
(50)
WA-11
(183)
OH-20
(311/247)
SC-8
(142)
OK-7
(36)
CT-7
(57)
MN-10
(193)
NV-5
(316/227)
GA-15
(134)
AK-3
(29)
VT-3
(60)
NM-5
(198)
WV-5
(321/222)
KY-8
(119)
ND-3
(26)
IL-21
(81)
MI-17
(215/340)
MO-11
(332/217)
LA-9
(111)
NE-5
(23)
CA-55
(136)
NH-4
(219/323)
AR-6
(338/206)
SD-3
(102)
ID-4
(18)
MD-10
(146)
OR-7
(226/319)
VA-13
(351/200)
IN-11
(99)
MS-6
(14)
NJ-15
(161)
PA-21
(247/312)
TN-11
(187)
KS-6
(88)
WY-3
(8)
DE-3
(164)
IA-7
(254/291)
AZ-10
(176)
MT-3
(82)
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, Bush won all the states up to and including New Hampshire (all Kerry's toss up states, but Michigan), he would have 323 electoral votes. Both candidates numbers are only totaled through their rival's toss up states. In those cases, Kerry's number is on the left and Bush's is on the right in italics.

***Florida is the state where Kerry crosses (or Bush would cross) the 270 electoral vote threshold to win the presidential election. That state is referred to as the victory line
.

As always, these presidential races come down to what happens in those states competitive enough to be considered swing states. In mid-August of 2004, John Kerry needed every last one of those states in shades of blue and Florida to even hypothetically cross over 270 electoral votes. 2008 and 2004 differ on the spectrum in the fact that in 2008, Barack Obama's toss up states push the partisan line beyond the victory line. The Illinois senator has two states in Nevada and Ohio that he could cede to McCain and still top 270 electoral votes. Kerry's lead in 2004 was much more tenuous. The junior senator from Massachusetts didn't have a similar cushion. In fact, four years ago, the partisan line and the victory line would have converged on Florida. All Bush had to do then was to swing the Sunshine state a little less than a percentage point and the election would have been his. As it turned out, he just had to show up in the state during a hurricane season that ravaged the Florida coast to accomplish that. Sure, both Ohio and Nevada were close enough to have been put on the mid-August Watch List in 2004. And Bush certainly had to work to keep the Buckeye state from turning blue.

The Watch List*
StateSwitch
Arizonafrom Bush lean
to Toss Up Bush
Floridafrom Toss Up Kerry
to Toss Up Bush
Michiganfrom Toss Up Kerryto Kerry lean
Minnesotafrom Kerry leanto Toss Up Kerry
Nevadafrom Toss Up Bush
to Toss Up Kerry
New Hampshire
from Toss Up Kerryto Kerry lean
New Mexico
from Kerry leanto Toss Up Kerry
Ohiofrom Toss Up Bush
to Toss Up Kerry
South Carolina
from Strong Bush
to Bush lean
Tennessee
from Bush lean
to Toss Up Bush
*Weighted Average within a fraction of a point of changing categories.

As you look at the Watch List, you'll note that seven of the ten states were on lines that would have put them close to switching in Kerry's direction. Part II will show that those shift didn't come to fruition for Kerry and that something entirely different happened between August and November 2004.


Recent Posts:
The Electoral College Spectrum

The Electoral College Map (8/10/08)

On VP Announcement Timing and Graphic Naming -- Some Housekeeping