Election 2012 Preview Number One: Obama 275, Romney 263

It has been three years since Barack Obama's near-landslide victory of 2008.  The question of whether he will be the seventh of the last nine elected Presidents to secure a second term (Bush 43, Clinton, Reagan, Nixon, Eisenhower, Roosevelt) or the third of the last nine to be a one-termer (Bush 41, Carter) is a close one.  Despite his fairly low approval ratings, Obama is apparently roughly a coinflip to win re-election, with the outcome hinging upon his opponent and whether the economy ticks up modestly in the next year.  This is a first in an irregular series of previews framing that contest and making observations about particular states.

Obama, Romney Splitting the Vote in the Gallup Twelve

When you win a near-landslide, as did Barack Obama in 2008, you have some margin of popular votes (53-45) and electoral votes (365-173) to spare.  Gallup.com has set about polling 12 potential 2012 swing states that Barack Obama won in 2008 (Nevada, Colorado, New Mexico, Iowa, Wisconsin, Michigan, Ohio, Florida, North Carolina, Virginia, Pennsylvania, and New Hampshire), which comprise a pool, of 151 electoral votes.  Gallup has correctly excluded Indiana from the swing state category.  Starting from the Obama 2008 states, adjusting for modest changes in the distribution of electoral votes among states based upon the 2010 census, and moving Indiana back into the red column, the tally would be Obama 347-191.  Accordingly, to get to the 270 electoral votes needed for re-election, the President must win 74 of the 151 "swing" EVs, yielding no more than 77 to the eventual Republican nominee, which this essay assumes will be Mitt Romney, as that is the most likely answer by far after FOTM (flavor of the month) Herman Cain's recent troubles.

Of course, when it comes to Obama's re-election prospects, it's all gloom and doom if you want it to be.  While actual news outlets like the Los Angeles Times correctly convey that Gallup is finding Obama and Romney even in a set of twelve swing states, while propaganda outlets like NewsMax suggest that the Gallup data as to Obama consist of the respondents' view that, by 60-37, they are not better off than three years, which it presents as dooming Obama.

My rough cut for today takes seriously the Gallup finding that currently, President Obama has an approval rating of 42% and a disapproval rating of 50%, but takes as equally important that when pitted against former Massachusetts Governor Mitt Romney, President Obama polls 47-47 nationally, and loses 47-46 among the twelve Gallup swing states.  Polling almost exactly evenly both nationally and among the swing states suggests that the EVs of the swing states will be split roughly down the middle.  I presently predict, in descending order of safety for Obama, that he will win:  (1) Michigan [16 EVs]; (2) Nevada [6 EVs]; (3) Pennsylvania [20 EVs]; (4) Colorado [9 EVs]; (5) North Carolina [15 EVs]; and (6) Virginia [13 EVs].  I presently predict that Romney, in descending order of safety, will win:  (1) New Hampshire [4 EVs]; (2) Ohio [18 EVs]; (3) Florida [29 EVs]; (4) Iowa [6 EVs]; (5) New Mexico [5 EVs]; and (6) Wisconsin [10 EVs].

Why This Split Among the Gallup Twelve

It is an interesting challenge to try to parse the data within Gallup's polls to predict state outcomes.  Gallup makes the point that in the twelve swing states, 59% of Republicans describe themselves as enthusiastic about voting, while only 48% of Democrats do.  Fine.  But there are two highly countervailing considerations.  

First, though more Republicans describe themselves as highly motivated to vote, there remain in poll internals roughly one-quarter of Republicans who either vote for Obama or claim they will vote for some third candidate, while Obama holds above 85% of self-described Democrats, leaking less of his base.  As I have written before, I attribute Romney's lower level of support within his base to a combination of evangelical conservative discomfort with Romney's Mormonism and Tea Party dislike for Romney.  I'm more a demographer in modeling things, so I think the former is more important.  But for Romney to only hold 74% or so of Republicans in trial heats with Obama is a major problem (though not necessarily insurmountable) for him, and it suggests that the meme that Republican voters remain "more" enthusiastic is overly simplistic.  More GOPers are highly motivated, and more GOPers seem like Romney refuseniks at this point.  This issue may cure up if and after Romney is nominated, much as Hillary Clinton fans came home to Barack Obama.  I tend to think some of these voters will come home, but fewer in proportion than Clinton primary voters who came home to Obama.  This will be a critical question for the general election.

Second, simply stating that Republicans are more motivated to vote by the significant but modest margin of 59-48% only describes the likelihood that a person will of their own volition make sure they vote.  But the 2008 election was greatly different from recent Presidential elections in that the Democratic candidate deployed a fierce ground game built around great data collection (as to supporters and potential supporters), David Plouffe's relentless focus on Obama's ground game, and an unprecedented fundraising advantage.  The results were lopsided wins in Nevada (12 percent) and Colorado (8.5 percent), and narrow wins in GOP Presidential bastions like Virginia and North Carolina, built around ground game and lopsided support from substantial populations of African-Americans.  Obama's fundraising radically outstrips Romney's at this point, and Romney is going to be spending at least through February fighting for the Republican nomination.  While third party groups are likely to spend more on indirect support of Romney, the sensible prediction is that Obama will have a better ground game, recourse to his 2008 data set, and more money to work from more offices with more volunteers, which will effectively equalize the enthusiasm advantage Romney would otherwise arguably enjoy.  It seems plausible after seeing the Democratic ground games overperform poll predictions in Senate races in Nevada, Colorado, and Pennsylvania to think that the ground game advantage with Obama on the ballot will be at least like that in 2010, if not that in 2008.  As such, I am tending to predict state-by-state based on 2008 results less the general national deviation for Obama, and not adjusting further down based upon Gallup's assessment that Republicans are generically more enthusiastic.

The countervailing considerations aside, four of the six states seem particularly sound in the fundamentals for Obama.  Michigan went so decisively for Obama in the general election (16.5 percent) that it is hard to imagine it coming all the way back to the Republican column.  Additionally, the auto bailout will help Obama with his base there.  This state has not gone Republican since 1984.  Nevada was a twelve percent win for Obama (55-43), and Harry Reid's resounding five percent victory in 2010 when polls showed him losing by three testify to the force of the Latino vote, the labor vote, and the Democratic machine in Clark County.  Colorado is a younger, Obama electorate state that trends more neoliberal and technophile, and it rejected Ken Buck for Michael Bennet in a generally dismal 2010 cycle.  It is hard to see Obama losing a state won thus in 2010.  Also, Nevada and Colorado are states with early voting and no-excuse absentee voting, a factor that works well with Obama's expected organizational advantage.  Pennsylvania is tougher.  Joe Sestak very narrowly lost the 2010 Senate race there (though he overperformed polls), and the state has the oldest electorate in the U.S., which isn't good for Obama.  Nonetheless, Obama won Pennsylvania by 10.4% (above his national win margin) by working the state as heavily as any campaign has ever worked it, while the McCain campaign devoted more time and attention to it than many potential pickups.  Like Michigan, Pennsylvania hasn't gone Republican since 1984, but here the skew toward Obama among younger voters hurts. 

In this model, two states necessary for Obama's re-election are Virginia and North Carolina, two states won on the strength of his redefinition of the electorate, turning out younger voters in greater numbers and winning them decisively while winning previously unseen margins among African-American voters, who also turned out in historically high numbers.  Siting the Democratic National Convention is Charlotte may help slightly, and as a banking center, Charlotte has done better in the recession than some major cities.  Obama's approval rating remains close to 50% there, above the current national trend.  In Virginia, Obama has been crushing Rick Perry in trial heats and performing well against Romney, suggesting that there will be some carryover from 2008 as well.  These states are virtual jump balls, but their large African-American voting populations, and slightly better economic performance during the current recession should make Obama's numbers drop considerably less than the national trend.

And why would Romney win roughly half of the twelve Gallup swing states?  Again, if the vote is split almost evenly among them, the current assumption from Gallup, they should be split in some nearly even fashion.  But New Hampshire is easy:  Obama is relatively unpopular there, Republicans won there as recently as 2000, Romney has a house there, and is likely to crush that primary out of the park, giving the state a positive feel for Romney as it sends him south as a frontrunner.  Ohio and Florida were wipeout states for Democrats in 2010, and Obama won them both by only 3 while winning nationally by 7.  In races roughly splitting the national popular vote, Florida and Ohio cannot plausibly supply the margin to get a Democrat to 270, 2000 notwithstanding; Kerry also underperformed his national trend in Florida, Obama even more so while winning it, and again, it trends older.  George W. Bush won Iowa once, and just as Iowa Democrats were very energized for Barack Obama after his win there, it is likely Iowa's evangelical Republican base will be more energized after the coming caucus, though this state is really a coinflip.  New Mexico too is roughly even, though outside Albuquerque, voters came home to the GOP in 2010 in New Mexico, drawing my nod.  

Wisconsin is troubling for Democrats, because it was such a narrow win in 2000 and 2004, and went so decisively against Russ Feingold in 2010, and also elected Scott Walker, and still has (even after two recall successes for Wisconsin antiWalkerites) a bare Republican majority in that state's Senate.  Obama's approval rating is higher than the national average in Wisconsin, which is a good omen, but the fact remains that the Democratic Party in Wisconsin is more liberal than the average state Democratic Party, and Obama should perform better in a state like North Carolina with a bigger black population and more suburban neoliberals (Raleigh-Durham) than in Wisconsin, with more white left-liberals and trade unionists.  

Are There States Outside the Gallup Twelve That Obama or Romney Could Likely Pick Off?

Yes.  Obama's campaign is talking up several:  Georgia and Arizona in particular.  Georgia is interesting, because Obama's approval rating there was most recently 48%, well above his national average.  I predicted incorrectly that Obama would win Georgia last time around, and he lost it by 5%.  But he also didn't work the state in media until the last week, and he had no ground game.  Obama's strong performance in the eyes of the center and right electorate on national security issues may buy him a few points with white voters there, and the math in Georgia does not require Obama to get above 30% of white voters to win.  Upper 20s works.  Given Georgia's large black population, and the evangelical discomfort with Mitt Romney's Mormonism, Georgia could be a pickup for Obama, though I can't see it being the state to get him to 270, which I think would be either New Mexico, Wisconsin, or North Carolina.  Arizona is likewise a state in which Obama can get close, but which would be an implausible pickup unless he wins the national popular vote by 2-3%.  True, one recent survey showed Obama ahead of Romney by a margin of 45-40, but that survey showed Obama winning Arizona independents roughly 50-30.  The questions are whether Arizona's independents are ready to swing back to the center after handing Republicans a supermajority in the statehouse and victories for every statewide office in 2010.  Notably, Latinos went fairly heavily for Democrat Terry Goddard in the 2010 Arizona gubernatorial election -- either 71% or 85% of the Latino vote (depending on who you ask, Pew or VotoLatino), but Goddard still lost 54-42.  Obama's approval rating in Arizona is simply too low to make this equation work, in my opinion, but if Obama tops 50% nationally, Arizona is definitely in play.  For the same reasons Georgia could be a pickup, look for Romney to win but struggle in South Carolina and Mississippi.

Romney could plausibly win formerly blue states presently seen as outside the Gallup frame of swing states.  President Obama's popularity in Oregon is not as high as in Washington, and Republicans have fared reasonably well in recent elections there given the state's financial difficulties in the recession.  Minnesota is historically a reasonably close state in Presidential elections.  Picking Tim Pawlenty as an attempt to reach into the upper Midwest (and as a bridge to evangelicals) could help make Romney competitive in the Horrible Mosquito State.  While independent-minded Maine has voted for the Democrat in each of the last five Presidential elections (and went for Obama by a 17 percent margin), it also has two Republican Senators.  Romney is from neighboring Massachusetts (where he was Governor) and New Hampshire (where he maintains a home), and is arguably one of the last of the breed of New England Republicans that once performed well throughout the region.  These are states Romney could make a push in.

Conclusion:  From Now To March Will Largely Frame the November 2012 Election

The suggestion that horserace focus is premature is silly.  An incumbent President's popularirty in the twelfth (now) and thirteenth quarters of their presidency has been highly predictive of reelection.  Also, there are fault lines presently exposed in Romney's profile that suggest openings for Obama:  (1) Romney's failure even now to run as well as the generic Republican, who beats Obama in most polling; (2) Romney's failure to poll as highly as Obama's disapprove rating, while Obama runs higher than his approve number against Romney, which suggests that voters who disapprove of both may pick Obama in significant numbers; and (3) Romney's weakness with the Republican base, both evidenced by his failure to move past Herman Cain even during Cain's atrocious week, and by the 25% or more of self-described Republicans who don't pick Romney in an Obama-Romney trial heat.  The challenge for Obama is simply to get some traction in job generation numbers, which would likely be enough to nudge him ahead.  For Romney, it is securing his own base, and getting independents excited about him.  The next five months won't end the campaign, but they will be pivotal in addressing all of these questions.