Monday, November 03, 2008

Turners

I’d like you to take a mental walk with me, to explore the Electoral map from the historical perspective. The reason for it will become self-evident, I believe.

The race to the White House is actually 51 separate races, each of which awards between 1 and 55 electoral votes (the 1 comes from Nebraska and Maine, which award individual electoral votes according to results from the state’s precincts. As Mister Gore found out in 2000, it is entirely possible for a candidate to lose the popular vote yet win the election, although because of the way states line up it is almost impossible for Obama to win the election without also winning the popular vote. Anyway, the finish line is set at 270 electoral votes, and the goal therefore is to reach or pass that line.

We’ve heard so much for so long about “blue” states and “red” states, that we start to miss the significance of those tags. First off, some states get tagged “red” or “blue” just because they went one way in the last election. But in some cases, the name is valid. What I want to do here, is separate the states which do not shift much, from the ones which truly are possible losses or gains, what I call “turners”. Any state can be had, of course, under the right circumstances, but the history gives us a good look at how likely that really is to happen.

First, let’s look at the true blue states, states which have generally gone democrat in presidential elections since 1948 and we should expect no different now. Looking at the election results, we see the following:

“Locks” for Obama:

District of Columbia: Since 1960 D 51%+ 100.0%, R 51%+ 0.0%, Democrat last 12 elections, avg since 1960 83.2-13.8 D, 89-09 D in 2004. For McCain to win, would require a 92-point swing from 2004. [92.0% chance Obama]

Hawaii: Since 1960 D 51%+ 58.3%, R 51%+ 16.7%, Democrat last 5 elections, avg since 1960 53.2-43.0 D, 54-45 D in 2004. For McCain to win would require an 11-point swing from 2004. [73.4% chance Obama]

Massachusetts: Since 1948 D 51%+ 66.7%, R 51%+ 20.0%, Democrat last 5 elections, avg since 1948 55.1-48.2 R, 62-37 D in 2004. For McCain to win would require an 27-point swing from 2004. [70.1% chance Obama]

Minnesota: Since 1948 D 51%+ 53.3%, R 51%+ 20.0%, Democrat last 8 elections, avg since 1948 47.8-44.7 R, 51-48 D in 2004. For McCain to win would require a 5-point swing from 2004. [70.5% chance Obama]


“Locks” for McCain:

Alaska: Since 1960 D 51%+ 8.3%, R 51%+ 75.0%, Republican last 10 elections, avg since 1960 56.0-40.7 R, 61-36 R in 2004. For Obama to win would require a 27-point swing from 2004. [80.0% chance McCain]

Arizona: Since 1948 D 51%+ 6.7%, R 51%+ 80.0%, Republican last 2 elections, avg since 1948 54.8-40.6 R, 55-44 R in 2004. For Obama to win would require a 12-point swing from 2004. [82.8% chance McCain]

Colorado: Since 1948 D 51%+ 13.3%, R 51%+ 73.3%, Republican last 3 elections, avg since 1948 52.3-42.7 R, 52-47 R in 2004. For Obama to win would require a 7-point swing from 2004. [75.0% chance McCain]

Florida: Since 1948 D 51%+ 13.3%, R 51%+ 53.3%, Republican last 2 elections, avg since 1948 51.5-42.9 R, 52-47 R in 2004. For Obama to win would require a 7-point swing from 2004. [70.6% chance McCain]

Idaho: Since 1948 D 51%+ 6.7%, R 51%+ 80.0%, Republican last 10 elections, avg since 1948 59.1-34.7 R, 68-30 R in 2004. For Obama to win would require a 40-point swing from 2004. [77.7% chance McCain]

Indiana: Since 1948 D 51%+ 6.7%, R 51%+ 60.0%, Republican last 10 elections, avg since 1948 54.8-41.6 R, 60-40 R in 2004. For Obama to win would require a 22-point swing from 2004. [80.9% chance McCain]

Kansas: Since 1948 D 51%+ 6.7%, R 51%+ 86.7%, Republican last 10 elections, avg since 1948 57.5-37.8 R, 62-37 R in 2004. For Obama to win would require a 27-point swing from 2004. [81.7% chance McCain]

Montana: Since 1948 D 51%+ 13.3%, R 51%+ 73.3%, Republican last 3 elections, avg since 1948 52.7-42.5 R, 59-39 R in 2004. For Obama to win would require a 22-point swing from 2004. [73.6% chance McCain]

Nebraska: Since 1948 D 51%+ 6.7%, R 51%+ 86.7%, Republican last 10 elections, avg since 1948 60.9-35.2 R, 66-33 R in 2004. For Obama to win would require a 35-point swing from 2004. [80.9% chance McCain]

New Hampshire: Since 1948 D 51%+ 6.7%, R 51%+ 66.7%, Democrat last 1 election, avg since 1948 53.5-42.3 R, 50-49 D in 2004. For McCain to win would require a 2-point swing from 2004. [75.5% chance McCain]

North Dakota: Since 1948 D 51%+ 6.7%, R 51%+ 80.0%, Republican last 10 elections, avg since 1948 56.8-38.3 R, 63-35 R in 2004. For Obama to win would require a 30-point swing from 2004. [81.2% chance McCain]

Ohio: Since 1948 D 51%+ 6.7%, R 51%+ 53.3%, Republican last 2 elections, avg since 1948 50.5-45.1 R, 50-46 R in 2004. For Obama to win would require a 6-point swing from 2004. [71.7% chance McCain]

Oklahoma: Since 1948 D 51%+ 13.3%, R 51%+ 60.3%, Republican last 10 elections, avg since 1948 55.1-40.7 R, 66-34 R in 2004. For Obama to win would require a 34-point swing from 2004. [74.0% chance McCain]

South Dakota: Since 1948 D 51%+ 6.7%, R 51%+ 73.3%, Republican last 10 elections, avg since 1948 54.9-41.5 R, 60-38 R in 2004. For Obama to win would require a 24-point swing from 2004. [81.5% chance McCain]

Utah: Since 1948 D 51%+ 13.3%, R 51%+ 80.0%, Republican last 10 elections, avg since 1948 60.5-34.4 R, 72-26 R in 2004. For Obama to win would require a 48-point swing from 2004. [74.2% chance McCain]

Virginia: Since 1948 D 51%+ 6.7%, R 51%+ 60.0%, Republican last 10 elections, avg since 1948 52.3-42.1 R, 54-45 R in 2004. For Obama to win would require an 11-point swing from 2004. [79.4% chance McCain]

Wyoming: Since 1948 D 51%+ 13.3%, R 51%+ 73.3%, Republican last 10 elections, avg since 1948 58.3-37.2 R, 69-29 R in 2004. For Obama to win would require a 42-point swing from 2004. [74.4% chance McCain]


States leaning Obama

California: Since 1948 D 51%+ 26.7%, R 51%+ 40.0%, Democrat last 4 elections, avg since 1948 48.1-47.3 R, 54-44 D in 2004. For McCain to win, would require an 11-point swing from 2004. [56.2% chance Obama]

Connecticut: Since 1948 D 51%+ 33.3%, R 51%+ 40.0%, Democrat last 4 elections, avg since 1948 47.9-47.9 tie, 54-44 D in 2004. For McCain to win, would require an 11-point swing from 2004. [55.0% chance Obama]

Delaware: Since 1948 D 51%+ 40.0%, R 51%+ 33.3%, Democrat last 4 elections, avg since 1948 48.0-47.9 D, 53-45 D in 2004. For McCain to win, would require a 10-point swing from 2004. [54.7% chance Obama]

Illinois: Since 1948 D 51%+ 26.7%, R 51%+ 33.3%, Democrat last 4 elections, avg since 1948 48.5-48.3 R, 55-45 D in 2004. For McCain to win would require a 12-point swing from 2004. [54.4% chance Obama]

Maryland: Since 1948 D 51%+ 40.0%, R 51%+ 33.3%, Democrat last 4 elections, avg since 1948 49.6-46.8 D, 56-43 D in 2004. For McCain to win would require a 15-point swing from 2004. [57.2% chance Obama]

New York: Since 1948 D 51%+ 46.7%, R 51%+ 26.7%, Democrat last 5 elections, avg since 1948 50.9-45.5 D, 58-40 D in 2004. For McCain to win would require a 19-point swing from 2004. [57.7% chance Obama]

Rhode Island: Since 1948 D 51%+ 53.3%, R 51%+ 26.7%, Democrat last 5 elections, avg since 1948 54.8-39.6 D, 59-39 D in 2004. For McCain to win would require a 22-point swing from 2004. [61.3% chance Obama]

Vermont: Since 1948 D 51%+ 26.7%, R 51%+ 60.0%, Democrat last 4 elections, avg since 1948 50.9-44.2 D, 59-39 D in 2004. For McCain to win would require a 22-point swing from 2004. [66.2% chance Obama]

Washington: Since 1948 D 51%+ 46.7%, R 51%+ 20.0%, Democrat last 5 elections, avg since 1948 50.9-45.3 D, 53-46 D in 2004. For McCain to win would require a 9-point swing from 2004. [69.1% chance Obama]


States leaning McCain:

Alabama: Since 1948 D 51%+ 26.7%, R 51%+ 46.7%, Republican last 7 elections, avg since 1948 56.8-33.5 R, 63-37 R in 2004. For Obama to win would require a 28-point swing from 2004. [56.8% chance McCain]

Arkansas: Since 1948 D 51%+ 46.7%, R 51%+ 33.3%, Republican last 2 elections, avg since 1948 48.6-45.1 D, 54-44 R in 2004. For Obama to win would require a 12-point swing from 2004. [55.4% chance McCain]

Georgia: Since 1948 D 51%+ 40.0%, R 51%+ 40.0%, Republican last 3 elections, avg since 1948 49.0-45.7 D, 57-41 R in 2004. For Obama to win would require an 18-point swing from 2004. [54.7% chance McCain]

Iowa: Since 1948 D 51%+ 13.3%, R 51%+ 46.7%, Republican last 1 election, avg since 1948 50.1-46.3 R, 50-49 R in 2004. For Obama to win would require a 3-point swing from 2004. [67.4% chance McCain]

Kentucky: Since 1948 D 51%+ 20.0%, R 51%+ 46.7%, Republican last 2 elections, avg since 1948 50.4-46.1 R, 60-40 R in 2004. For Obama to win would require a 22-point swing from 2004. [62.5% chance McCain]

Louisiana: Since 1948 D 51%+ 20.0%, R 51%+ 53.3%, Republican last 2 elections, avg since 1948 46.4-42.6 R, 57-42 R in 2004. For Obama to win would require a 17-point swing from 2004. [58.8% chance McCain]

Mississippi: Since 1948 D 51%+ 13.3%, R 51%+ 40.0%, Republican last 7 elections, avg since 1948 47.2-37.3 R, 59-40 R in 2004. For Obama to win would require a 20-point swing from 2004. [61.9% chance McCain]

Missouri: Since 1948 D 51%+ 20.0%, R 51%+ 73.3%, Republican last 2 elections, avg since 1948 49.1-48.1 R, 53-46 R in 2004. For Obama to win would require a 9-point swing from 2004. [59.8% chance McCain]

Nevada: Since 1948 D 51%+ 13.3%, R 51%+ 40.0%, Republican last 2 elections, avg since 1948 52.3-42.4 R, 50-48 R in 2004. For Obama to win would require a 3-point swing from 2004. [69.0% chance McCain]

New Jersey: Since 1948 D 51%+ 26.7%, R 51%+ 40.0%, Democrat last 4 elections, avg since 1948 49.7-46.3 R, 53-46 D in 2004. For McCain to win would require an 8-point swing from 2004. [56.2% chance McCain]

New Mexico: Since 1948 D 51%+ 20.0%, R 51%+ 53.3%, Republican last 1 election, avg since 1948 50.3-46.2 R, 50-49 R in 2004. For Obama to win would require a 2-point swing from 2004. [62.8% chance McCain]

South Carolina: Since 1948 D 51%+ 20.0%, R 51%+ 40.0%, Republican last 7 elections, avg since 1948 48.4-41.5 R, 58-41 R in 2004. For Obama to win would require a 19-point swing from 2004. [61.5% chance McCain]

Tennessee: Since 1948 D 51%+ 13.3%, R 51%+ 33.3%, Republican last 2 elections, avg since 1948 49.6-45.4 R, 57-43 R in 2004. For Obama to win would require a 16-point swing from 2004. [66.3% chance McCain]

Texas: Since 1948 D 51%+ 26.7%, R 51%+ 46.7%, Republican last 7 elections, avg since 1948 50.5-44.9 R, 61-38 R in 2004. For Obama to win would require a 25-point swing from 2004. [59.7% chance McCain]

West Virginia: Since 1948 D 51%+ 26.7%, R 51%+ 46.7%, Republican last 2 elections, avg since 1948 48.3-47.1 R, 56-43 R in 2004. For Obama to win would require a 15-point swing from 2004. [58.1% chance McCain]


Toss-up States:

Maine: Since 1948 D 51%+ 26.7%, R 51%+ 40.0%, Democrat last 4 elections, avg since 1948 49.8-45.5 R, 56-43 D in 2004. For McCain to win would require a 25-point swing from 2004. [51.6% chance Obama]

Michigan: Since 1948 D 51%+ 33.3%, R 51%+ 40.0%, Democrat last 4 elections, avg since 1948 48.2-47.8 R, 51-48 D in 2004. For McCain to win would require a 5-point swing from 2004. [53.3% chance McCain]

North Carolina: Since 1948 D 51%+ 40.0%, R 51%+ 33.3%, Republican last 7 elections, avg since 1948 49.8-45.7 R, 56-44 R in 2004. For Obama to win would require a 14-point swing from 2004. [51.6% chance Obama]

Oregon: Since 1948 D 51%+ 13.3%, R 51%+ 33.3%, Democrat last 5 elections, avg since 1948 53.3-46.7 R, 51-47 D in 2004. For McCain to win would require a 6-point swing from 2004. [50.2% chance Obama]

Pennsylvania: Since 1948 D 51%+ 26.7%, R 51%+ 33.3%, Democrat last 4 elections, avg since 1948 48.8-48.0 D, 51-48 D in 2004. For McCain to win would require a 5-point swing from 2004. [51.5% chance Obama]

Wisconsin: Since 1948 D 51%+ 20.0%, R 51%+ 33.3%, Democrat last 5 elections, avg since 1948 48.7-46.9 R, 50-49 D in 2004. For McCain to win would require a 2-point swing from 2004. [52.3% chance McCain]


The percentage chance of a candidate taking a state is a formula incorporating the percentage of wins by a party in a state since 1948, the percentage of elections where a party candidate claims 51% or more of the vote, the average support for a party in a state since 1948, the lowest and highest support levels for a party in a state since 1948, the RCP average polling for each candidate, and the 2004 results by party.

Please note that these are historical patterns only, and do not take into account demographic changes. But it does lend some historical perspective on the situation. Take it as you will.

7 comments:

Anonymous said...

Interesting, DJ. Here's how I think it will go down: exit polls show Obama way up--this is crap due to selection bias, but meant to suppress Western vote. First returns could show Mac up or very close in PA and VA. MSM will deny this is trend, FOX will hesitate. If PA goes for Mac, we got a race. If VA goes (assuming NC goes Mac and it will), Obamaphiles panic. FL becomes a must but Mac will probably hold it. So..OH is the ballgame and most likely litigation state. Obama will have lawyers file to keep polls open longer to allow more ACORNs to get to polls in big cities. GOP lawyers will try to challenge potentially fradulent voters. If OH within 3% and dispositive, we will be a long time in deciding this election. Whadya think?

OBloodyHell said...

> If PA goes for Mac, we got a race.

Heh. If PA goes for McCain, he's won it right there. That's a state which Obama has supposedly had in his pocket for months. And it means that people have clued into Obama's BS and voted against it. And if they're doing it there, they'll do it in a lot more places.

Huan said...

thank you for your blog analysis
win or lose


i have questions regarding the polls

1. who pays for the public polls like Rasmussen, Zogby, Gallup & Pew? (i.e. follow the money)

2. could the polls be presented as say 51 Obama and 45 McCain with a MoE being 3% be statistically true though the original mean/median was 48 & 48. or even 51 McCain and 45 Obama. I mean that if the observed value was 48, then the truth could be anywhere between 51 and 46, no?
What do you think the chances are that the poll could be manipulated to such and extent and yet be published in such a way as to remain statistically true but clearly misleading?


again, thanks .

Anonymous said...

Huan,
The answer is yes. Check out DJ's post called Yahoo, Indeed.
http://stolenthunder.blogspot.com/2008/10/yahoo-indeed.html

Anonymous said...

DJ, What do you think of the logic of the Marston Chronicles which today predicts a 310 to 228 McCain/Palin landslide? See his website at http://www.marstonchronicles.info/index.php?option=com_content&task=view&id=95&Itemid=119
I am wondering how his take on the impact of PUMA's on polling error correlates to your own analysis.

Anonymous said...

DJ-
hey, I want McCain as much as anyone. However, while the history is important, I am not sure this really tells me much. The electorate has changed so much, the party principles have changed dramatically, and then there has been this major realignment in how states vote starting with Reagan. I just dont think the analysis is valid with all these changes that have occurred.

That said, I do believe McCain will win and that the current batch of polls do not reflect a truly randomized sample of the population.

Rick E

Unknown said...

I'm also optimistic about McCain's chances, and I do think there is abundant evidence that many of these "respected" polling firms have unrealistic models. However, I think that the way these states are ranked isn't realistic.

For example, you have Alabama ranked something near a toss-up state. Alabama would secede before it went for Obama. The probability it goes for McCain should be around 90%.

Now I don't have a well-thought out mathematical model to guide me, but my guess is the electoral map will be very close to 2004, with no single state deviating more than 5 percentage points from the results of 2004.

I just don't see people completely flipping their ideology in one election cycle. That's why I always thought it was ridiculous when I saw states like Indiana showing up as battleground states that were tight in the polls.

Bush won that state by 21 percentage points in 2004. States just don't change that rapidly.