Sunday, April 22, 2012

The 2012 Presidential Map: Structural Advantage - Obama, Why Did Americans Elect Fizzle?

Almost Two Weeks Into the General, Obama Holds an Electoral College Lead

As we have now effectively entered the general election phase of the 2012 contest, we are getting real red meat for those of us who love analyzing polls and trends.  We have 9 new national polls within the past week and state-level polling in 42 of the 50 states, including virtually all of the states that could reasonably be considered contestable (North Dakota being the only state that is conceivably contestable where there isn't a poll and even North Dakota isn't likely to be close unless Obama wins by quite a hefty margin.)

Crunching through the numbers, I see two trends emerging:

(1) Abnormally Large Poll Divergance
The CNN/Opinion Research poll had President Obama leading Mitt Romney nationally by 9%.  The NBC News/Wall Street Journal poll had the race at 6% lead for Obama.  Both the Rasmussen and Gallup Tracking Polls have Romney in the lead, albeit by only 1%.

What explains this large poll divergence?  There could be multiple reasons.

The first obvious reason would be the sample the pollsters are attempting to represent.  Rasmussen targets Likely Voters, using a voter screen to determine who among the polled is likely to actually show up on election day, whereas all of the other polls are targeted Registered Voters, a self-selected group that may or may not vote.  Virtually all polls move to a Likely Voter model close to election day, but most start with Registered Voters early in the season when it is typically very hard to project turnout.  But Rasmussen had it at Romney +1%, the same total as Gallup, who is using Registered Voters.

The second possible reason is that the race isn't very well decided at this early stage.  Despite the long primary season, the truth is that most general election swing voters haven't been closely watching the Republican primaries and debates and probably don't know that much about Mitt Romney yet.  In this situation, small changes in the news cycle and poll timing can have outsized effects on the results.

The third is sample size.  The Rasmussen, Gallup and Quinnipiac polls have large sample sizes (1500, 2200 and 2577 respectively) and have a much tighter clustering of poll data at Romney +1, Romney +1 and Obama +4, where all of the other polls have sample sizes of 910 or less.

At any rate, utilizing my weighting methodology, which considers sample size in the weighting of the polls, at this point I have Obama at +0.9% in the aggregate of all the polling, a lead, but certainly a very small one (for reference, in the two closest elections in recent history, Al Gore's popular vote margin over George W. Bush in 2000 was 0.5% and Bush's margin over John Kerry was 2.4%.)

(2) Structural Advantage - Obama
It is very early to make this kind of projection, but the state-level polling data indicate that the way the demographics are breaking, President Obama has a structural advantage in the electoral college in a close race.  With a less than 1% lead nationally, the President leads in many of the key swing states, including New Hampshire, Ohio, Virginia, Iowa, Pennsylvania, Nevada, Colorado, Minnesota and Michigan.

As things stand today, I project a 303-235 electoral college lead for President Obama, in spite of being narrowly behind in the State of Florida.

The map is below (created with the help of our friends at

For the state-by-state numbers, I have utilized a combination of state-level and national polling data, depending on the number of state-level polls available.  For states where no polling is available, I've used the 2008 election results as a baseline, modified for the difference between current national polls and the 2008 national results.

In terms of the contested states, here is where we stand:
Safe Romney (51 Electoral Votes)
Wyoming, Utah, Oklahoma, Idaho, Alaska, Alabama,  Arkansas, Louisiana, West Virginia

Strong Romney (116 Electoral Votes)
Mississippi, Nebraska, Kentucky, North Dakota, Kansas, Tennessee, Texas, Georgia, South Carolina, South Dakota, Arizona

Likely Romney (14 Electoral Votes)
Montana, Indiana

Lean Romney (54 Electoral Votes)
North Carolina, Missouri, Florida

Lean Obama (41 Electoral Votes)
New Hampshire, Ohio, Virginia, Iowa

Likely Obama (70 Electoral Votes)
Pennsylvania, Nevada, Minnesota, Colorado, Michigan, Wisconsin

Strong Obama (97 Electoral Votes)
New Jersey, Oregon, Washington, New Mexico, Connecticut, Maine, Illinois, Massachusetts, Delaware, Maryland, Rhode Island

Safe Obama (95 Electoral Votes)
California, New York, Vermont, Hawaii, District of Columbia

Of worthy note, there are presently no states that John McCain won in 2008 that are meaningfully in play for 2012.  Missouri is the closest at a 4.1% lead for Romney, which makes sense, if you recall that John McCain won it by the narrowest of margins in 2008.  Montana and Arizona are the next closest at 9.2% and 10.2% leads for Romney respectively, which puts them only on the fringe of being contestable.

Meanwhile, Romney, as things stand today, would pick up 3 Obama states from 2008: Indiana, North Carolina and Florida.  Obama won Indiana and North Carolina by very narrow margins in 2008 and Florida, while he won decisively, he won by significantly less than his national margin, making it a logical flip in a close race.

But Romney would need to do much more to pick up the Presidency.  The easiest path for the winning 270 Electoral Vote total for him to win both Virginia and Ohio (both presently lean Obama states) and 1 of either New Hampshire or Iowa (winning just New Hampshire would give him 270, winning just Iowa would give him 272, winning both would boost his total to 276.)

Alternatively, he could take John McCain's (unsuccessful) strategy from 2008 and swing for the fences in Pennsylvania.  If he won his current states plus PA, he would have 255 electoral votes as a base, meaning that just winning Ohio would put him over the top (at 273) or combining a win in Virginia with a smaller state such as New Hampshire (Pennsylvania, Virginia plus New Hampshire would give him 272 electoral votes.)

Clearly, Florida, where Romney leads by 0.1% as things stand today, is critically important to his election bid.  It's virtually impossible to draw a feasible map for Romney that doesn't include Florida's 28 electoral votes.  Give him Florida and he has all those options in the mid-west.  Take Florida away and you could give Romney New Hampshire, Ohio, Virginia, Iowa AND Pennsylvania and he still falls short with only 268 Electoral Votes.

So, in my rank order of importance to the race, considering the closeness of states, their electoral vote contribution and their relative impact on the possible scenarios, I think the most hotly contested states will be as follows:
1. Florida
2. Ohio
3. Virginia
4. Pennsylvania
5. North Carolina

If Obama wins Florida (a toss-up) or North Carolina (a tougher haul, but he did it in 2008), it is virtually impossible for Romney to win.  If Romney takes both, the other three states become huge in deciding the outcome.

Below is the state-by-state spread as things stand today.

Third Parties Falter Again in America
American's Elect sounded like such a great concept.  A well-funded organization with the mission of bringing a qualified third party candidate to the American voters, using a web-based, democratic process to bring a true choice to America.  American's Elect was well-enough backed that they appear likely to get on the ballot in all 50 states and constructed a slick website and good marketing campaign, attracting press from major national news organizations such as CNN.

You'd think with the seeming hatred most American hold for both our major political parties that 2012 would be fertile ground for such an effort.  Gridlock, partisanship and public disapproval for elected officials are at generational highs.

Yet, American's Elect is flopping big time.  A check of their website shows that two weeks from their online convention, no declared candidate has even reached 4,000 supporters, FAR less than the 50,000 spread over 10 states that the site had originally set as a bar for candidates.  And the candidates are less than enthralling, with the leaders being Buddy Roemer (3,639 supporters), an ex-Governor who couldn't even get on the stage when he tried to run as a Republican and Rocky Anderson (2,073 supporters), a former liberal mayor who has never even been a Congressman, Senator or Governor.  Hardly the resumes of highly-qualified centrist candidates that Americans Elect had hoped to recruit, such as Michael Bloomberg or Jon Huntsman. 

It strikes me that while we all talk of the need for a third-party, we aren't particularly serious.

If you like this site, tell your friends.

No comments: