Swing State Clusters Tell Story of Potential "Map-Changing" Obama Landslide

by: Paul Rosenberg

Sat Aug 02, 2008 at 13:03


This map from Pollster.com--like the maps Chris has been posting here for more than two months--tells a vivid story of a potential Obama landslide in the Electoral College, a "map-changing" victory, if you will:

Of course it's only "potential" at this point.  There's a long way to go till election day.  But as I noted in a quick hit yesterday, also over at Pollster.com, UMass Amherst Poly Sci Prof Brian Schaffner explained:that "Obama had been tied or ahead in 50 consecutive national polls through Sunday," and that meant there was only a .0000000000000009 chance that the race was actually tied at this point.

Schaffner lead his brief post with this priceless quote:

"We have a race that by every measure of every poll is a statistical dead heat. McCain's not supposed to be in this thing, and Obama's supposed to be blowing everybody away and it just isn't happening, at least to this point."

Lou Dobbs (July 17th, Lou Dobbs Tonight)

The state-level view only serves to underscore the boundless cluelessness of Versailles.

In this diary, I want to do a brief walk-through of four regions where three or more states are in some state of play, each of which alone would spell big trouble for John McCain.  You can see them quite easily on the map above: The "Mountain Swing" region consisting of the three yellow states Nevada, Arizona and Colorado, plus the light blue state of New Mexico; the "Northern Plains Swing" region consisting of yellow states Montana and North Dakota, plus the light red state of South Dakota; the "Midwest Swing" region, consisting of yellow states Missouri and Indiana, plus light blue states Iowa and Ohio; and the "Southeast Swing" region, consisting of yellow states Virginia, North Carolina and Florida. Put them all together, and they don't just spell "trouble" with capital T and that rhymes with P and that stands for poll.  They spell unmitigated disaster.  Sort of like the last 8 years.

In a follow-up diary, I'm going to look at the "mapping-changing" meme in terms of election results from 1896 to date. So think of this as the microscopic view, with the telescopic view to follow.
Paul Rosenberg :: Swing State Clusters Tell Story of Potential "Map-Changing" Obama Landslide
Mountain Swing/Battlegrounds

The Mountain Swing region is sufficient in itself to win the election for Obama, if he also holds the 2004 Kerry states.  Of course, winning Arizona would be deeply embarrasing to Senator McCain, and even though it's actually a possibility, it's not likely to be a priority--even though Clinton did win Arizona in 1996.  Even without Arizona, however, the 19 EVs of the other three states plus Iowa are enough for Obama.  Clinton won Nevada and New Mexico both times he ran, and won Colorado in 1992, in addition to Arizona in 1996, so winning at least three of these states would be one vital factor in changing the 2008 map to look more like the Clinton maps of the 1990s than the Bush maps of the 2000s.

Let's turn to the individual states.

Nevada

Nevada poll trends:

Chris:   Obama: 42.3 / McCain: 43.0 - Margin: 0.7 McCain
Pollster.com:   Obama: 41.3 / McCain: 44.9 - Margin: 3.6 McCain.
538:    Obama: 44.7 / McCain: 45.0 - Margin: 0.3 McCain

Clinton won here twice. With Yucca Mountain, and growing strength of union and Latino voters, we should be doing better here.  We should win this one.


Arizona

Arizona poll trends:

Chris:   Obama: 37.3 / McCain: 44.0 -- Margin: 6.7% McCain
Pollster.com:   Obama: 40.3 / McCain: 43.8 -- Margin: 3.5 McCain
538:   Obama: 39.3  / McCain: 47.2 - Margin: 7.9 McCain

It's McCain's home state, but Clinton won here once, and a lot of Arizona Republicans don't like McCain.  Nor do many of Arizona's reporters.  Pollster.com shows it much closer than others, and there's no reason to expect a big Obama push here, but it is worth noting that if the election depended on it, Obama could conceivably win here.  And if this is a landslide realigning election, he could well win here anyway, totally shocking any pundits who notice.


Colorado

Colorado poll trends:

Chris:   Obama: 45.3 / McCain: 43.5 -- Margin: 1.8% Obama
Pollster.com:   Obama: 45.8 / McCain: 44.1 -- Margin: 1.7 Obama
538:   Obama: 46.7 / McCain:  44.3  - Margin: 2.5 Obama

The margins aren't big, but Colorado looks very good on the Senate level, and a vigorous GOTV operation should make this Obama's to lose.


New Mexico

New Mexico poll trends:

Chris:   Obama: 48.5 / McCain: 40.3 -- Margin: 8.2 Obama
Pollster.com:   Obama: 49.2 / McCain: 41.0 -- Margin: 8.2 Obama
538:   Obama: 49.1 / McCain:  43.0 - Margin: 6.1 Obama

The beginning of the end of New Mexico as a swing state?  Could be.


Northern Plains Swing/Battlegrounds

At three votes each, Montana, North Dakota and South Dakota aren't exactly Electoral College powerhouses.  This is the one swing region that can't win the election for Obama.  But if he wins even one state here, it's almost certain that he'll be doing well nationwide, and should have nothing to worry about.  What's more, the psychological significance of these states punches way above their weight after eight long years of clueless yammering about "Red States" vs. "Blue States".  Solid leads in any of these states in the last week or two of the election should be deeply demoralizing to the McCain campaign.

South Dakota may seem out of reach now, but the other two are clearly in play, and a strong Democratic tide could put South Dakota in play as well. If that happens, all bets are off.  Clinton won Montana in 1992, so the recent state-level resurgence there is matched by some historical grounds for hope at the presidential level this year.  But North Dakota and South Dakota haven't gone Democratic since Johnson's 1964 landslide, and last time before that was FDR's massive landslides of 1932 and 1936, so their impact would be truly momentous.

Montana

Montana poll trends:

Chris:   Obama: 43.3 / McCain: 46.0 -- Margin: 2.7 McCain
Pollster.com:    -- No Trend
538:   Obama: 46.7 / McCain:  46.2 - Margin: 0.5 Obama

Clinton won here once, and the locals have been doing a bang-up job, so the state seems definitely in play.


North Dakota

North Dakota poll trends:

Chris:   Obama: 42.3 / McCain: 43.5 -- Margin: 1.2 McCain
Pollster.com:   -- No Trend
538:   ND Obama: 44.4 / McCain:  46.2 - Margin: 1.8 McCain

After wining it twice in his first landslide elections, FDR lost it in 1940, and the Democrats have only won here once since-in LBJ's 1964 landslide-so the fact that this state is in play is one of the clearest signs there is that we're likely headed for a 1932-style election.


South Dakota

South Dakota poll trends:

Chris:   Obama: 38.8 / McCain: 47.5 -- Margin: 8.7 McCain
Pollster.com:   -- No Trend
538:   Obama: 43.8 / McCain:  47.5  - Margin: 3.7 McCain

It's not looking as good as North Dakota, but historically, the two have voted in lockstep for so long, it's not out of the question.


Midwest Swing/Battlegrounds

With recent battlegound states Iowa and Ohio looking fairly strong for Obama (Iowa moreso), the emergence of Missouri and Indiana as "toss-ups" to be fought over is one of the key indicators that we may be seeing widespread map changes this election.  Obama can win the election simply by winning Iowa (where he's lead consistently) and Ohio (where he has lead in most polls since May) in addition to the 2004 Kerry states. But a lot more could be possible.

Clinton won Missouri both times out, but aside from that, Carter '76 was the only time Democrats have won Missouri since LBJ's 1964 landslide.  Although geographically part of the Midwest, its Border State status tends to be more salient-only Southern Democrats have carried it since 1964, so Obama winning here would be huge.  But Indiana would be even huger, as no Democrat has carried the state since LBJ's landslide in 1964, and before that FDR's back-to-back landslides in 1932 and 1936.  Given that history, the mere fact that these two states are even in play marks an historic occasion.

Missouri

Missouri poll trends:

Chris:   Obama: 43.8 / McCain: 45.4 -- Margin: 1.6 McCain
Pollster.com:   Obama: 43.8 / McCain: 44.6 -- Margin: 0.7 McCain
538:   Obama: 46.1 / McCain:   46.6  - Margin: 0.6 McCain

This is a key Clinton state that Obama definitely could win back.  He's putting a lot of campaign staff into the state, while McCain has been advertising.  This should be a state to watch in the last two months.


Iowa

Iowa poll trends:

Chris:   Obama: 46.0 / McCain: 39.8 -- Margin: 6.2 Obama
Pollster.com:   Obama: 45.2 / McCain: 38.7 -- Margin: 6.5 Obama
538:   Obama: 49.4 / McCain:  41.3 - Margin: 8.1 Obama

Not even close this time?  It seems quite possible now.


Ohio

Ohio poll trends:

Chris:   Obama: 45.4 / McCain: 44.0 -- Margin: 1.4 Obama
Pollster.com:   Obama: 45.9 / McCain: 42.4 -- Margin: 3.5 Obama
538:   Obama:  46.8 / McCain:  44.1  - Margin: 2.7 Obama

Not as solid as one might hope, but the state-level Democratic tide from 2006 seems to be holding. A strong GOTV effort in the state could help pick up House seats that eluded our grasp in 2006 as well.


Indiana

Indiana poll trends:

Chris:   Obama: 43.3 / McCain: 45.3 -- Margin: 2.0 McCain
Pollster.com:   Obama: 42.4 / McCain: 43.0 -- Margin: 0.6 McCain
538:   Obama: 46.7 / McCain:   46.3 - Margin: 0.3 Obama

Indiana?  This was the Klan's playground in the 1920s, and is the GOP's favorite Northern state.  The fact that it's competitive now is another key indicator that a 1932-style realigning election is distinctly possible.


Southeast Swing/Battlegrounds

In the 1990s, Clinton did well for himself in the South.  But he did not do well where Obama is showing strength-Virginia, North Carolina and Florida.   Here in the Southeast, Obama can win the election by supplementing the 2004 Kerry states with either Florida alone or Virginia and North Carolina together.  All three, and you have the makings of a blowout, as he would almost certainly make large gains elsewhere as well.

Look backwards, Clinton won Florida in '96 and Georgia in '92, and aside from that he was shut out of the lowland Southern states from Virginia to Mississippi.  His strength lay in the Appalachian states of West Virginia, Kentucky and Tennessee, and the Mississippi River states of Louisiana and Arkansas, connecting to the border state of Missouri.  Carter carried all the South except Virginia in 1976, and his home state of Georgia in 1980, but the last time the Democrats carried the trifecta of Virginia, North Carolina and Florida was Johnson's 1964 landslide, and last time before that was Truman in 1948.  These are not the Deep South states that Goldwater took in 1964 or that Strom Thurmond and the Dixiecrats won in 1948.  But they're hardly easy pickings for Democrats.

Virginia

Virginia poll trends:

Chris:   Obama: 46.8 / McCain: 44.8 -- Margin: 2.0 Obama
Pollster.com:   Obama: 46.7 / McCain: 44.1 -- Margin: 2.6 Obama
538:   Obama: 47.2 / McCain:  45.6  - Margin: 1.6 Obama

Virginia has been a long time coming for Democrats, and it looks like it's finally arrived.


North Carolina

North Carolina poll trends:

Chris:   Obama: 43.5 / McCain: 47.0 -- Margin: 3.5 McCain
Pollster.com:   Obama: 44.0 / McCain: 46.0 / Barr: 2.0 -- Margin: 2.0 McCain
538:   Obama: 44,1 / McCain:  46.8 - Margin: 2.7 McCain

Virginia was expected to be in contention, regardless of who the Democratic nominee was.  North Carolina is a gift-and another indicator that this election could be something special.


Florida

Florida poll trends:

Chris:   Obama: 46.8 / McCain: 46.3 -- Margin: 0.5 Obama
Pollster.com:   Obama: 44.9 / McCain: 45.7 -- Margin: 0.8 McCain
538:   Obama: 45.6 / McCain:   46.4 - Margin: 0.8 McCain

For some reason (I wonder why?) Democrats don't seem to like to talk alot about Florida, but the numbers show it to be pretty tight.  The last two months could be quite a ride-or we could see early signs of an Obama blowout.


Summary

The opportunities listed above are just that-opportunities.  But if we put them all together, and look at the best-case scenario, it's clear that a major change in the electoral map is clearly possible for the Democrats, while even the most optomistic Republican scenario would be lucky to pick up one or two seats more than Bush.

Here is the best-case scenario map based on all the above:

That's quite a change from the 2000/2004 map.

And here's the best-case scenario for the Republicans:

Really not much different from the 2000/2004 map.

Yet another demonstration that change is not the GOP's friend.

But is that actually the best Democrats could do?  In a word, no.  Using my "Edwards Bump" analysis from early July, which produced this map:

Regardless of whether Edwards is the VP, this map shows where the strongest possible ticket could pick up additional states--or at least put them into play.  Using this as our guide, we can project a conceivable Democratic landslide as large as the following:

Of course what is conceivable is nowhere near guaranteed.  But given the constricted nature of the Versailles-dominated horse-race discourse, it is a healthy addition to step back and consider the possibility of truly dramatic change.  This diary has looked at how that might shape up, from a regionally-based look at where the (mostly-) new battleground states are.  Next up will be a look at map-changing in an historical perspective, from 1896 to date.

Tags: , , , , (All Tags)
Print Friendly View Send As Email
Swing State Clusters Tell Story of Potential "Map-Changing" Obama Landslide | 22 comments
One question (4.00 / 1)
I hate to disturb all of that thorough analysis with a stupid question, but I am wondering about this part:

"Prof Brian Schaffner explained: that 'Obama had been tied or ahead in 50 consecutive national polls through Sunday,' and that meant there was only a .0000000000000009 chance that the race was actually tied at this point."

Here's my question: How can we use polls from last week, last month and three months ago (which I assume is where the 50 consecutive national polls begin) as a way of assessing the probability that Obama is ahead TODAY?

In other words, the probability argument would make lots of sense to me if the polls were taken around the same period, or if we could assume that opinions have remained static.  But the former assumption is not true, and how can we trust the latter assumption?

I have the feeling I'm missing something basic here.

Knock Off Some Zeroes (0.00 / 0)
Schaffner's argument was about the ongoing narrative, in which case it makes perfect sense to consider the entire time period.

For a narrower look at whether polls are tied now, one would have to cut down the time-period, say to a week or two.  But the same logic would apply,  If the race really were tied, then there'd be roughly as many McCain-leading polls as there are Obama-leading ones.... and there aren't.

While you've made a valid distinction, and one we should keep in mind, your observation is also quite likely to used to shore up past mistakes (not that that is your fault!) along the lines of "Well, we were wrong before, but now we're right."  However, this is not a compelling argument without a sober assessment of why you were wrong before and a responsible re-evalutation.

The day Versailles does that, I'm going ice-skating in Hell.

"Senate passes expanded GI bill despite Bush, McCain opposition"

[ Parent ]
Good point (0.00 / 0)
I assume the figure was arrived at by simply multiplying the odds of each poll being wrong, but that seems like an illicit move to try to establish the odds that the race is tied right now. If you had 50 polls come out today, and they all said Obama was ahead, then this would be legitimate.

But as it stands, all that Schaffner has really proved are the odds that the race has been tied all along - or, the odds that the polling has repeatedly, over time, miscast Obama as the leader. But surely no one is suggesting that.  

[ Parent ]
"surely no one is suggesting that" (0.00 / 0)
Well, every week someone has claimed that the race is tied.  So I'd say, yes, they do claim that.


New Jersey politics at Blue Jersey.
[ Parent ]
Not the same thing (0.00 / 0)
And in fact, that gets to the heart of why this "stat" is misleading. Is there anyone who is claiming that Obama is now and always has been tied/trailing? Not just one person this week, another person next week, etc. - is anyone actually saying that Obama has never at any point been ahead? I mean, maybe someone is saying that, but it's just so incredibly dumb that I guess I was hoping otherwise.

With a caveat: it's only dumb if the argument is based on raw statistics/margin of error stuff. If, on the other hand, you believe that the polls are systematically making Obama look stronger than he really is - e.g. because of some kind of Wilder effect, or something - that's a totally different issue. Such an argument may be just as misguided, I don't know, but it wouldn't be disproved by the professor's analysis.

[ Parent ]
That's Not The Way It Works (0.00 / 0)
It's not a question of one person saying the same thing over and over again.  It's a question of Versailles as a whole saying the same thing over and over again.

So no one's responsible, see?  

"Senate passes expanded GI bill despite Bush, McCain opposition"

[ Parent ]
A race is not a snapshot (4.00 / 1)
So if Obama has been tied or ahead in 50 consecutive national polls, then when the Talking Heads have talked about a "statistical tie" they are either stupid or doing what they are told to do, to belittle Obama's chances, lest a bandwagon effect set in.

Or to put it another way, Obama has been tied or ahead for three months. So when Repubs wish that some poll will show him behind TODAY, they are wishing for something that they have been wishing for and not getting for three months, but that doesn't make it any more likely to be true at long last as of TODAY.

[ Parent ]
New day (0.00 / 0)
What a relief to wake up to this terrific post, after yesterday's (perfectly justifiable) communal obsession with McCain's Trash Talk Express.  
Well, It Does Help Provide Some Context For McCain's Panic, No? (0.00 / 0)
[ Parent ]
Yes (0.00 / 0)
The charts do indeed contextualize and explain McCain's panic. He has lost his bearings because he doesn't like where he is right now. Of course, this same tactic of exoticizing the opponent, combined with electoral theft, worked for Bush, so Obama and his supporters need to stay savvy.  
[ Parent ]
Number of places (4.00 / 1)
It's amazing to consider the amount of states that Obama can choose to play offense in.  You really have to hope the GOTV machine works as well for the campaign in the general as it did in the primaries.

Even as an Edwards supporter, however, I'm very skeptical that a VP could put that many states into play.  I think a good VP selection would be more likely to boost Obama in states where he is already very competitive and put him over the top in those states, rather than add new battleground states to the map.

But Edwards DID Help In Places Like Virginia (0.00 / 0)
And, in fact, he helps precisely with some of those demographics where Obama needs it most.  

Plus, being a White Southerner further adds to the credibility that Edwards could bring South Carolina, Georgia and Mississippi over to the Dems.

It's certainly possible that Clark could have a similar impact, possibly putting Arkansas into the Dem column.

However, this was clearly just the icing on the cake.

"Senate passes expanded GI bill despite Bush, McCain opposition"

[ Parent ]
Isn't Edwards kind of a nonstarter at this point though? n/t (4.00 / 1)

[ Parent ]
Not Really The Point (0.00 / 0)
During my whole series of posts on Edwards I was never assuming that Obama would choose him.  In fact, I think Obama's judgement is extremely poor on such major matters.  The analysis was using Edwards to make a point, and that point is still valid--there's a good deal more electoral opportunity out there than Obama on his own can manifest at this point in time.

"Senate passes expanded GI bill despite Bush, McCain opposition"
[ Parent ]
one quibble (4.00 / 1)
I would include SD and one of Nebraska's EV's as in the best case scenario.

Other than that, this looks about right.

I love the idea of averaging Chris', 538's, and pollster's margins into one average.  

SD, Yes (4.00 / 1)
I can't remember what was going through my head that I didn't include it.  The Nebraska EV is a good point, but it's not consistent with the methodology I was employing, so I'm comfortable not including it, though I should have made this caveat clear.

It should be obvious that a "best case scenario" can always be exceeded by unexpected good fortune.  It's just not something one should count on.

You play to win each game.  But if the other team forfeits one by not showing up, and you get the win, while resting your team, well, how can something like that not exceed your "best case scenario"?

"Senate passes expanded GI bill despite Bush, McCain opposition"

[ Parent ]
More suggestive evidence (0.00 / 0)
Is Kentucky a Midwestern state like Indiana, Ohio, and Missouri? To stick with one pollster, Rasmussen saw McSame ahead by 25 points on May 22, but that lead had shrunk to 16 points by June 25 and only 9 points on July 29. Other polls give widely varying numbers, but the trend here is striking. I don't expect to see Obama closing the gap and taking the state, because this isn't one where he's investing in his vaunted ground operation or doing any campaigning -- though some TV signals must flow easily across the Ohio River from the adjoining states where he is on the air.

West Virginia is another Border State that's part Midwestern, part Appalachian. Almost forgotten and almost unpolled since the primary where Obama basically took a pass, its one poll, back in June, showed McSame ahead by a measly 8 points.

If we can pick up Ohio, Indiana, and Missouri in this election, surely Kentucky and West Virginia will be battleground states in 2012.

Next Cycle's Battleground States (0.00 / 0)
The big picture pattern--which I'll discuss as part of my next diary--is that realigning elections are followed by even bigger wins.  This was the case even in 1936, following an almost total sweep in 1932.

So the opportunities you're pointing to would be right in line with historical precednt, to dovetail with the poll-based argument you've presented.

The Midwest

The term "Midwest" is somewhat ambiguous (as are virtually all regional terms), which is readily visible on the map at the Midwest Wikipedia entry.  Missouri, Kentucky and West Virginia are all part of the outer Midwest, along with the Plains states from Kansas to North Dakota.

OTOH, in "With God On Our Side," Bob Dylan refered to the Midwest as "the country I come from," in which case one has to have clearly defined borders, no?

"Senate passes expanded GI bill despite Bush, McCain opposition"

[ Parent ]
Regions (4.00 / 1)
Remember when Time-Life Books published series with a set of volumes eventually giving COMPLETE coverage of a subject?

A series about the US sliced and diced the country into a 12-volume series, and truly, the more regions, the more coherent the definitions. New England and the Pacific Coast States are clearly defined regions; the South or the Midwest less so.

IIRC that Time-Life series defined the Midwest as Ohio, Indiana, Illinois, Michigan, and Wisconsin, and I think of those as the core of the Midwest.

Anyway, my point about Kentucky, West Virginia, and Missouri is that they are indeed Border States, bordering the Midwest and sharing some important characteristics, but not others. WVa used to have steel mills and unionized coal mines that gave it much in common with the industrial and unionized core states of the Midwest. Kentucky has manufacturing, like the Toyota factory and one making GE appliances, that give it something in common with Michigan. The two big cities in Missouri act like Midwestern cities, but much of the rest of the state is Ozark-Appalachian.

But if the core Midwest states are strongly solidifying as blue, it's to be expected that these partly Midwestern Border States will feel -- and show -- that effect as well.

[ Parent ]
Good Point (0.00 / 0)
I think it also makes sense to look at the role that rising hope can play.  If hope is fulfilled--not necessarily totally, but enough to make effort worthwhile--then this can spur further engagement, and a dynamic of increased progressive activism, voting and governance can follow.  But disappointment can have the opposite effect.  And disappointment while others nearby are experiencing progress can be even more devastating.

Here, I'm thinking very much of the dynamic as Ohio tries to right itself, while Kentucky and West Virginia are right next door.

"Senate passes expanded GI bill despite Bush, McCain opposition"

[ Parent ]
EV counts for "best case" maps? (0.00 / 0)
Paul,
A very interesting and digestible analysis.  Thanks.
Did I miss something, or did you not include total electoral vote counts for the various "best case" scenario maps?  Do you have these handy?
No I Didn't (0.00 / 0)
Here they are:

Obama best case (w/o VP assist):

Obama:  391
McCain: 147

Obama best case (w/ VP assist):

Obama: 454
McCain: 84

McCain best case:

Obama:  231
McCain: 307


"Senate passes expanded GI bill despite Bush, McCain opposition"

[ Parent ]
Swing State Clusters Tell Story of Potential "Map-Changing" Obama Landslide | 22 comments
Donate to Open Left









QUICK HITS

Friends of the Earth thanks the OpenLeft community for the ideas you generate and your contributions to the progressive movement.

blog advertising is good for you
blog advertising is good for you
LIBERAL INSTITUTIONS
STATE BLOGS
SEARCH

   

Advanced Search

Home | RSS/XML Feed | Register | Contact

  © OpenLeft 2007-2009, Powered by SoapBlox