Sunday, July 27, 2008

The Electoral College Map (7/27/08)

Well, there's a lot to go over today. The latter half of the week saw a break from the 10 poll norm we had settled into over the course of the last few electoral college breakdowns with a typical amount of polling from Rasmussen (six new polls) in addition to another four swing state polling release from Quinnipiac. In total 17 new polls from 15 states were released since Wednesday. As I have mentioned in previous weeks, there seems to have been a relative settling down of FHQ's weighted averages while at the same time the polls have been fairly volatile (A state like Minnesota comes to mind. Results out of the North Star state have ranged anywhere from a 2 to 18 point lead for Obama recently.). And that pattern is continued with this new polling.

New Polls (July 23-26)
StatePollMargin
(With Leaners/ Without Leaners)
California
Rasmussen
+10/+12
Colorado
Quinnipiac
+2
Colorado
Frederick
+4
Florida
Rasmussen
+2/+1
Maine
Critical Insights
+20
Michigan
Quinnipiac
+4
Minnesota
Quinnipiac+2
Minnesota
Rasmussen+13/+12
Mississippi
Research 2000/Daily Kos+9
New Hampshire
Rasmussen
+4/+6
New Jersey
Gannett
+14
New Mexico
Rasmussen
+6/+5
North Dakota
Research 2000/Daily Kos+3
Pennsylvania
Rasmussen
+6/+5
South Carolina
Research 2000/Daily Kos
+13
Virginia
Public Policy Polling
+2
Wisconsin
Quinnipiac
+11

One thing you will notice that is different about the New Polls table is that all the information from Rasmussen is included: both "with leaners" and "without leaners" results. And I should note that I'll continue to use the "with leaners" data to compile the information for our electoral college maps. Fret not, though. I plan on regularly updating the differences in those two numbers from Rasmussen and will chart those changes on a separate map as warranted. The map "with leaners" will continue to be the official FHQ electoral college map, though. As of now the only change from using either set of numbers continues to be Ohio. The map below (with leaners) shows the Buckeye state's 20 electoral votes going to McCain. When the leaners are excluded, Obama takes an equally narrow lead. That is the only case where a state's electoral votes change hands or categories.

...for now. Additional future polling could--and probably will--change that.

Changes (July 23-26)
StateBeforeAfter
South CarolinaToss Up McCain
McCain lean

All those polls and the only change is South Carolina moving more solidly toward McCain into the Arizona senator's lean category. The last time South Carolina shifted was following the Zogby poll that gave Obama a one point edge in the Palmetto state. As I said then, I didn't really think of South Carolina as or expect it to necessarily be an Obama state come early November. I said then, and I still maintain, that South Carolina is a state that Obama can improve upon past Democratic nominee's performances, but to win it would be a tremendous coup, not to mention an indication that Obama has won a landslide victory. Even when the Zogby poll is omitted, South Carolina remains a McCain lean. This is due in large part to the fact that this most recent Research 2000 poll in the Palmetto state is the only one to place the race outside of the single digit range. This poll, then, acts to pull the average here back into the range where we would expect South Carolina to be.
[Click Map to Enlarge]

Let's talk about landslides for a minute. An Obama victory that includes South Carolina may be considered a landslide but that characterization may be an exaggeration depending upon how landslide is defined. If you rank the states for each candidate (1 through 51) based on the states' current averages South Carolina would rank 34 on Obama's list. On the current map McCain has 26 states and 260 electoral votes while Obama holds leads in 24 states (and the District of Columbia) totalling 278 electoral votes. If South Carolina is the last state on the list that Obama is able to pick off from McCain (and assuming the Illinois senator wins the states between South Carolina and where his list of states ends), he would gain 101 electoral votes (That's a group of states that includes all thoes above shaded in pink and Alaska.) on the total in the map above. His advantage over McCain in that scenario would be 379-159, a margin comparable to Bill Clinton's 1992 electoral college win over the first president Bush (370-168). Was 1992 a landslide? Maybe. Maybe not. Compared to both Reagan blowouts or Nixon's thrashing of Mondale in 1972, that doesn't look so bad. But compared to the tight margins of the 2000 and 2004, it could be portrayed as such in November by the media. Is such a scenario likely? No, but I suppose it is possible.

The reality of the situation now is that this appears to be a pretty close election given the current state of the polls and where most handicappers have things pegged currently (Though this appears to be up for debate. See the discussion between Abramowitz, Mann and Sabato and Campbell.). Obama has a lead, but not an insurmountable one. By our calculations, that margin is but 18 electoral votes, 278-260. During the latter half of this past week, McCain shifted South Carolina's 8 electoral votes into his lean category. And while that brings the Arizona senator total between strong and lean states to 157, that number still falls shy of Obama's stong total of 175. It should be noted, however, that the Obama's lean and toss up electoral vote totals are both currently hovering around 50. The Illinois senator has a solid coalition of states put together, but there still may be work to do putting together a winning combination. All of those (both light blue and blue) states are state where McCain is challenging Obama.

The Watch List*
StateSwitch
Arizonafrom Strong McCainto McCain lean
Floridafrom Toss Up McCain
to McCain lean
Georgiafrom McCain leanto Strong McCain
Minnesotafrom Strong Obamato Obama lean
Mississippifrom McCain leanto Strong McCain
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 McCainto Toss Up Obama
Virginiafrom Toss Up McCainto Toss Up Obama
*Weighted Average within a fraction of a point of changing categories.

And the Watch List? Well, North Dakota, South Carolina and Wisconsin have exited. The first is firmly within the toss up category, but favors McCain enough that new (and negative to McCain) polling won't necessarily move North Dakota into Obama territory. South Carolina, after a brief stay in the toss up category, is now firmly within the McCain lean area. And Wisconsin edges above the cut off point, but solidly within the Obama lean category. New Mexico is the only addition to the list. The Land of Enchantment's weighted average has been around 6 points (similar to Wisconsin) for a while now. Recently, it has been just above six, but with the new poll (Rasmussen +6) comes in just under that line. And that puts it on the watch. Essentially, Wisconsin and New Mexico switched places.

I'll be back later today with an updated look at the recent Rasmussen numbers and some other stuff too.


Recent Posts:
Guam: Oh Well, So Much for Frontloading the General Election

A Game of Vice Presidential Chicken

Is Obama Getting a Bounce from His Trip Abroad? A Different Approach

2 comments:

Anonymous said...

As we know, July polls are often poor indicators of November results. However, I believe that the early state polls give us some sense of the relative strength of the candidates across states. The national electorate may move 5 or 10 points in either direction over the next 3 months. However, that potential may already be seen in Josh's electoral college map. Such a shift would move most or all of the tossup states (and maybe some of the leaning states) to the winning candidate. Thinking of the states ordered from most Democratic to most Republican (as Josh does in his discussion of South Carolina) can be illuminating. Of course the order is not yet set or fully known; there will be local or regional changes not related to national trends. Nonetheless, one can imagine a cutting line separating states won by Obama from states won by McCain. In addition, there would be a line representing an equal division of electoral votes. If the order of states is accurate, then whichever candidate wins the state through which this line passes wins the election. In 2000, it was Florida; in 2004, it was Ohio.

Anonymous said...

Good points, Paul.

What you are talking about as far as the spectrum of states is similar to what FiveThirtyEight is doing with their Tipping Point graphic (right side column on their site).

To flesh this out a little, though, let's look at how that looks with our average. To save space, I'll just focus on the states considered toss ups. And again, as you said, this is early and these rankings will be somewhat fluid.

The most Democratic toss ups are at the top with most Republican toss ups at the bottom.

21. MI 17
22. CO 9
23. NH 4
---------What McCain needs for 270
24. PA 21
25. NV 5
---------Current Line
26. OH 20
27. VA 13
28. ND 3
29. MT 3
30. IN 11
31. MO 11
32. FL 27
33. NC 15

If the election were held today (and if our measure is accurate), Obama would win. He could even spare Nevada and still win. McCain on the other hand would have to make inroads in Nevada (it is close enough as it is) and pick off Pennsylvania as well. Pennsylvania would push McCain past 270 to 286 electoral votes and victory. And that is interesting in light of the comments I made in like the second or third electoral college post (It was actually the fourth.) where Obama and McCain were exactly tied. Pennsylvania is the new Ohio is the new Florida. Pennsylvania is the tipping point here for now. Though, the way things are trending either Colorado or Michigan could slip in there and claim that distinction.