Saturday, October 27, 2012

10 Days Out - Electoral College Predictions - a Tossup Election?

We are now in the final sprint, today a look at the race and where the swing states are...

The Prof's Debate #3 Grades:

Obama   B
Romney  B

The third debate was a tie.  Foreign policy, while of utmost importance simply isn't what this particular election is about.  Romney did what he had to and appeared calm and presidential, while Obama scored on debating points.  Obama was ahead for much of the debate, but Romney closed well, therefore a tie in my book.

Let's look at the electoral landscape yet again...Bear in mind the polling is now all post-debates and is based on likely voter, not registered voter models.

Based on the site, the red and blue states indicate the "safe" states for each campaign with the swing states shown in beige.  Let's look at each swing state and see how things have changes from my September analysis.  My October analysis for each state is indicated in italics.

North Carolina - 15 EVs:  
The Tarheel State which is historically Republican was captured by Obama by a whisker in 2008.  This was mainly driven by high African-American and college student turnout. Current head to head polling shows Romney with a small, but consistent lead.
Early Prediction - Romney  

September: But some recent polling shows it now tied.  Unless there is a total meltdown, I still think Romney squeaks it out here.   
September Prediction - Romney
October: Looks like with Romney's October surge, that NC has been put away.  Early voting numbers show that Obama's advantage of  2008 has evaporated.  I think Romney wins comfortably in NC now. October Prediction - Romney

Nevada - 6 EVs: 
This state is an interesting mix with both a sizable Hispanic and Mormon population.  Won easily by Obama in 2008, Nevada is a battleground once more. However, I think that the president's advantage among Hispanics will pull it out for him in the end.
Early Prediction - Obama

September: Still agree with this analysis although the polls remain fairly close. September Prediction - Obama

October: Polls remain close, now a tossup, but I think early voting numbers are enough to favor a small Obama victory.                                                                                October Prediction - Obama

Iowa - 6 EVs:  
Recent polling shows this state as tied which is good news for Romney. Obama is well-organized here and this will be nip and tuck.  However, Iowa is very white and this is a group that has been deserting Obama in droves since 2008. Current polling shows a neck and neck race here. Both candidates are working very hard at this one.    
Early Prediction - Romney

September Obama is in better shape in Iowa now than he was a month ago and based on his momentum, I will flip Iowa to the Democrats...for now...                              September prediction - Obama

October: Polls have shifted this back to a tie.  Romney's strong lead among white voters nationwide will just put him over the top here methinks...                                           October Prediction - Romney

Colorado - 9 EVs:
This is an eclectic state with a conservative rural and military population combined with liberal meccas like Boulder.  A high Hispanic turnout should help Obama, but early polling shows a very close race.  I have to flip a coin on this one.  
Early Prediction - Obama

September: The coin I flipped in June is still on edge, but state polling is showing either a tie or small Obama lead so we will keep this the same as well. 
September Prediction - Obama

October: Romney has moved into a very small lead here and the early voting number appear to favor him so I am switching COP back to red. This is a close one.
October Predication - Romney

Florida - 29 EVs:  
This is a must-win for Romney and lends some weight to selecting Sen. Rubio for VP. Florida historically leans Republican, but Obama will be very competitive here.  But I see Romney pulling it out in the end based on his strong lead among voters over age 50.  
Early Prediction - Romney

September: Polls show an absolute tie in Florida.  Not good news for Romney who hoped to put away this swing state early. Without a win here, his path to 270 becomes almost impossible. No longer think Romney has the advantage here as he is being hammered over Medicare and immigration.  But with the state's economy in poor shape comparatively I will leave it where it was, but would not be surprised at all if it flips back to the president.  
September Prediction - Romney

October: Polling now shows that the nationwide Romney surge has hit the beach in FL and is eroding Obama's chances of keeping this state blue.  Feeling more confident that Romney steals this state's critical 29 electoral votes. 
October Prediction - Romney

Virginia - 13 EVs:  
The Old Dominion is getting more purple than it was thanks to an influx of wealthy, educated, and socially liberal whites in the DC suburbs, but much of it remains part of the old South.  Republicans won big here in 2010, but I think Obama prevails here for a second time due to a large and enthusiastic turnout in Fairfax County. This will be very closely contested.  
Early Prediction - Obama

September: Polling now shows Obama expanding his lead.  This is another must-win for Romney, but if things continue as they are I see him losing by at least 5 points. Obama really has solidified his lead in northern Virginia and the DC suburbs.  The addition of former VA Rep and Libertarian presidential candidate Virgil Goode also drains some votes that would likely go to Romney.  
September Prediction - Obama

October: As in several other swing states, Romney has taken a small lead in most polls. Obama will do well in Fairfax County and the DC suburbs, but Romney will sweep in coal country and likely among military voters in the Virginia Beach area.
October Prediction - Romney

Pennsylvania - 20 EVs:  
Described by strategist James Carville as Philadelphia and Pittsburgh on the ends with Alabama in the middle, PA is a state with a split personality.  Romney will win in the rural central part of the state and in coal country while Obama will rake in the urban areas.  The election will be won in suburbia but with a historic Democratic lean I will place PA (for now) as an Obama state.
Early Prediction - Obama

September: Polling shows this state leaning further into Obama's camp.  The fact that Romney is not campaigning or advertising heavily in PA is another indicator that he is writing this state off.  Obama should win by a comfortable margin, some polls now have him in a double digit lead.  
September Prediction - Obama

October: Polling has tightened again with a few showing a tie, but Obama seems to be maintaining a lead of between 4 and 6 points.  If Romney has a huge turnout in rural areas he could take it (and thus the election), but I think Obama hangs on here based on the overwhelming Democratic registration advantage. Also, Romney has not committed significant resources here.
October Prediction - Obama

Ohio - 18 EVs:   
Another must win for Romney; no Republican has won the presidency in recent history without the Buckeye state. Again this is a battle to be waged in suburban Cleveland, Columbus, and Cincinnati with Romney winning big in rural areas and Obama holding the cities.  This will be close, but I give the edge to Romney at this point.  
Early Prediction - Romney

September: Not any more.  Romney is consistently down by ever widening margins especially over the past couple of weeks.  Looks like the auto bailout and Ohio's comparatively good local economy is boosting the president.  In some polling he is winning women by 20 points and the race is even among men.  Romney needs to turn this one around immediately or he is in deep trouble.  
September Prediction - Obama

October: What a difference a month makes.  This is back to a coin toss.  Obama is maintaining a very slight advantage in the polls, but all are within the margin of error.  Ohio is indeed ground zero of this election - if Obama loses OH, he is a one term president.  If he wins, Romney's path to 270 is very difficult.  For now, I will keep it with Obama, but I can easily see Romney squeaking it out here.  May be a long night!
October Prediction - Obama

Michigan - 16 EVs:  
Obama seemed very strong here a month ago, but recent polling shows a closer race.  Romney has family ties here, but his opposition to the auto bailout has hurt him. MI has been trending Democratic in recent years and while close, I think Obama will squeak it out here. Early prediction - Obama

September: Like Ohio, the president seems to be pulling away.  Another state where Romney has stopped advertising as well. looks like another state Obama does not have to worry about or dedicate a lot of resources to.
September Prediction - Obama

October: Michigan, like PA has closed.  Unemployment remains high, thus Romney does have a shot here, but I am betting the auto bailout helps keep this state blue.
October Prediction - Obama

Wisconsin - 10 EVs:  
This state was very close in 2000 and 2004, but was won easily by Obama in 2008.  The recent failure of the recall election against Gov. Walker and a strong Republican organization makes this very competitive this time around.  A possible Paul Ryan VP pick makes this even more interesting. Another coin flip...
Early Prediction - Romney

September: The Ryan pick has made this state competitive, but polling in WI is showing the same trend of independents and women to Obama.  Ryan may not be enough here as Romney continues to falter, in fact, I am now thinking that Obama wins this by a few.
September Prediction - Obama

October: Like other swing states, the gap has closed back to a tie.  WI is getting a lot of attention as it is crucial for Romney if he loses OH.  Ryan is helping here and the momentum favors Romney.  I am going to move it back to red, but this is a 50/50 state.
October Prediction - Romney

New Hampshire - 4 EVs:  
In an election as potentially close as this one, the Granite State's four electoral voes may decide the winner. Historically Republican, but trending Democratic, this will be close.  Romney has spent a lot of time here, but polling shows that his move to the political right may hurt him in this libertarian state.  One more coin flip.
Early Prediction - Obama

September:Based on consistent polling showing an Obama advantage, I am keeping this where it is.
September Prediction - Obama

October: Polling has tightened here (like everywhere else!)  Some are showing Romney in the lead, but the socially conservative nature of the national GOP simply plays very poorly in libertarian NH and should keep Obama ahead on election night - barely.
October Prediction - Obama 

Based on these predictions, this is how the map would look...

If this holds, Romney wins the presidency 273 to 265 - about as razor thin a margin as one can get.

Bear in mind, I am assigning these tossup states based on polling, campaign activity, some early voting data, and what I think the likely voter model will be. If I am off just a bit, my predictions will be wrong.  It is not implausible that either candidate is able to have a solid Electoral College victory.

Why have so many states flipped - some of my reasoning...
  • Romney has taken a national polling lead of between 1 and 4 points depending on the poll.  National numbers matter insofar as they show trends that sometimes state polling does not pick up.  If Romney wins the national vote by 2-4 points, it becomes almost impossible to lose the Electoral College.
  • Romney is doing extremely well among white voters in general and suburban independents in particular.  Based on this, I am tilting Iowa and Wisconsin into his column as these are white and generally rural states.
  • Voters see the economy as the number one issue and Romney has an advantage on this issue.
  •  While Obama is still winning with personal popularity, the debates have helped Romney rehab his image and his favorables now outweigh his unfavorables.
  • Republican enthusiasm and likelihood to vote is sky high.
  • Romney looks more competitive in Pennsylvania and Michigan,.  While he may not win them, it compels Obama to divert campaign resources there...and away from other states.  For example, Romney's recent ad buy in Minnesota is probably a head fake, but again forces Obama on the defensive.
  • Obama is campaigning on the defensive and trying to hold the Ohio firewall - another sign that this is a very close race and that he has not regained the momentum - yet
Now, even though I have flipped my model to Romney, I still believe that Obama has the slight advantage in winning reelection. He has more paths to 270 than does Romney.  For example flip tossup state Colorado and Obama wins.

Reasons why Obama may win...
  • Overwhelming support among minority voters.  The nation is slowly becoming less white; if demographics is destiny this may be enough to eke out wins in some of the tossups.
  • Obama may have the better "ground game" that will turn out his voters, especially with so many early voters,.  Early voters were Obama's key to victory in 2008.
  • Romney is not doing as well among women as he needs to. The Democrats are working the abortion issue very hard and it may be enough to persuade younger unmarried women to vote.  This is a demographic that typically has low turnout.
  • Obama can point to the economic conditions starting to look up.  This stay the course argument may be enough to cut into Romney's advantage on this issue.
All in all, the Prof will be getting progressively less sleep as the day draws nigh.  I will publish final predictions with vote percentages in each state (yes, really out on a limb!) next weekend.  Also, expect a Warren-Brown update this week as well  - hint, ikt ain;t looking good for Brown!


The Prof