Obama’s Resilient New Coalition and Romney’s Prairie Fire Option

The available state and national polling suggests that Obama has maintained high levels of support among minorities and college educated voters, even while suffering historic losses among white voters without a college degree. The uneven changes in Obama’s coalition have clear consequences for the electoral college. In “new coalition” states like North Carolina, Obama has remained surprisingly competitive, while Obama is mired well beneath 50% in “old coalition” states like Ohio, a state that seems unlikely to support Obama in a close election.

Many analysts assume that Romney can recapture Virginia and Colorado in 2012, two states that offered support to President Bush in both 2000 and 2004. Karl Rove included Virginia as one of the three states in his “3-2-1″ electoral strategy. Certainly Romney can win either state if he prevails by a sufficient margin nationally, but I increasingly question whether the Romney campaign should count on either state in a relatively close national election. The state polling in Virginia and Colorado is sparse, but the polling and demographics suggest that both states will be very difficult for Romney to win in a close election, at least as long as Romney struggles to make gains among college educated voters.

Demographically, Obama’s coalitions in Virginia and Colorado ought to be among the most resilient in the country. Only 18% of Obama supporters in Virginia were whites without a college degree, less than any other competitive state, and Colorado was  second with 25%. Although Virginia was somewhat closer than Colorado, as Obama won by only 6.3% in 2008, Obama benefits from proportionately smaller dependence on white voters. Only 51% of Obama’s Virginia supporters were white in 2008, compared to 81% in Colorado, and the vast majority were African American. In other words, even though Obama won Virginia by less than Colorado, his base is stronger and there are proportionately fewer persuadable voters.

Mathematically, it is extremely challenging for Romney to win Virginia or Colorado without meaningful gains among college educated whites. In Virginia, Romney would need to reduce Obama’s share of the non-college white vote to just 24% to reduce Obama’s statewide tally to 49% – a 25% reduction. Similarly, in Colorado, Romney would need to reduce Obama’s share of the non-college white vote to just 30% – a 29% reduction. While these types of gains are possible, the Romney campaign would not be wise to bet on it.

The state polling is sparse but tends to confirm that Obama is well positioned in both states. Most polling in Virginia shows Obama leading with nearly 50% of the vote, and strong showings in polls of North Carolina, a similar state, provide added confidence. In Colorado, PPP is responsible for the majority of polls, but Obama is looking very strong in Colorado, even for PPP. State polling from elsewhere in the southwest, including Arizona, tends to confirm Obama’s standing.

Romney will struggle to win Virginia and Colorado as long as Romney struggles to make gains with college educated whites. Although such gains are certainly possible, there are reasons to doubt whether the Romney campaign can produce substantial improvements with college educated whites. Most fundamentally, these are voters who currently support and approve of Obama’s job performance, perhaps due to greater economic confidence and cultural issues. Winning over Obama supporters is clearly the path of most resistance.

Neither Virginia nor Colorado voted for Obama by an overwhelming margin in 2008, and if Obama was hemorrhaging support evenly across his diverse coalition, both would be prime GOP targets. However, Obama’s losses are far from balanced. Obama’s coalition has a narrow but deep wound among white working class voters. This wound is potentially fatal, but non-lethal in Colorado and Virginia – two states overwhelmingly composed of voters likely to continue to support Obama. Romney certainly can win these states, but he will be fighting on demographically unfriendly terrain.

Although it is too early to tell, there is a real possibility that these states will “lean” Obama in November, even in a close national election, if Romney can’t make gains among college educated whites. To compensate, the Romney campaign might be wise to consider an alternative but counter-intuitive strategy exploiting Obama’s weakness among white working class voters by targeting competitive states where Obama is most dependent on their support. In this regard, four states stand out: Iowa, New Hampshire, Wisconsin, and Minnesota.

Many may find it counter-intuitive to suggest that Romney might stand a better chance in these so-called “blue wall” states than Colorado and Virginia, but none should be considered out of reach for Republicans. Each of these states has given the Democrats less than 50% in a Presidential election since 1992, and Bush outright won Iowa and New Hampshire. Wisconsin only voted for Kerry and Gore by a combined 17,000 votes, so it’d be hard to characterize that state as anything like a wall.

Most importantly, Democrats are heavily dependent on white working class voters in all three states. More than 50% of Obama’s 2008 supporters in Iowa and Wisconsin were whites without college degrees, compared to 18% and 25% in Colorado and Virginia. New Hampshire and Minnesota were somewhat lower at 41% and 45%, but still much higher than the “new coalition” states.

While Obama won each of these states by more than Colorado or Virginia, Obama’s margin wasn’t always intimidating, especially considering the demographics. In Iowa, New Hampshire, and Minnesota, Obama won by less than 11% – only marginally more than Obama’s 9% margin in Colorado. Wisconsin was a borderline blowout at 14%, but Obama’s victory was built on presumably persuadable white Bush-Obama voters without a college degree. Indeed, a uniform and proportional swing among non-college white voters would give Romney victories in all four states before Colorado or Virginia.

While state-level polling is inconclusive, there is some evidence suggesting that these states may be more competitive. Recent polls in Wisconsin show a tight race, with Obama beneath 50%. While PPP and SurveyUSA, two polling firms showing Obama faring extremely well everywhere, are responsible for most polls in Minnesota and Iowa, the most recent poll from a different and highly respected firm showed Romney up 2% in Iowa. Similarly, internals on many recent national polls show Obama flagging in the Midwest (perhaps I’ll put this together later today).

Without more state polling, it is difficult to say whether any of these “blue wall” states are more competitive than Colorado or Virginia. National-level demographic trends need not manifest uniformly or proportionately across the swing states, and it is certainly possible that Obama might hold firm in the Midwest, even while losing more white working class voters elsewhere. Indeed, this has happened before. Over the last two decades, Democrats have lost millions of white working class voters across Appalachia, but seem to be in approximately the same spot in the Upper Midwest.

In the absence of better polling data, campaign activity can help identify competitive states. Although there is relatively little activity in Wisconsin and Minnesota, both campaigns are taking Iowa seriously. Romney recently described the threatening national debt as a “prairie fire” in Iowa and Romney’s first advertising buy included Iowa, not Colorado. If Romney thought Colorado was an easier win than Iowa, perhaps they would have gone with “forest fire.”

It is worth noting that a GOP win in Iowa would allow Romney to secure the Presidency without Colorado, although a win in Virginia would still be necessary, and New Hampshire required to avoid a 269-269 split, which actually looks quite plausible:

It seems relatively unlikely that Iowa could vote for Romney without moving Wisconsin into the toss-up column, as the two states have unsurprisingly tended to move and vote in unison. Even if the Romney campaign’s internal polling finds Wisconsin less competitive than they’d like, Romney’s gains in Iowa suggest that similar gains in Wisconsin could be forthcoming with additional effort. If Romney could win in both Iowa and Wisconsin, it would allow him to overcome a loss in Virginia, so long as he also carried New Hampshire.

Colorado and Virginia are competitive states, but it is not hard to envision how they could lean-Obama in November, even in a close national election, unless Romney makes larger inroads among college educated voters. The Romney campaign should keep their options open by adopting a ‘prairie fire’ strategy to vigorously contest states where Obama’s coalition is heavily dependent on white voters without a college degree, like Iowa, Wisconsin, Minnesota, and New Hampshire. There is no guarantee that national demographic trends allow Romney to win all of these states in a close election, but there isn’t a guarantee that Romney can make meaningful gains in the diverse and educated suburbs of Denver or Washington, either.