Last night, I had the good fortune to be part of a small gathering of reporters assembled by a quartet of top political scientists who have embarked on an effort to analyze voter opinion in the 2012 election at a level of depth and nuance beyond what we’ve managed in past years. A centerpiece of the effort is their attempt to gauge voter response to the ads that are already crowding the airwaves in battleground states. And their initial findings go a long way toward explaining why people in Richmond and Dayton and Denver are going to be seeing so much of Mitt Romney singing "America the Beautiful." 

Under the ad rating project—a coordinated effort between Vanderbilt University and the YouGov polling outfit—voters are shown presidential campaign ads and later asked for their responses to them and their candidate preference after viewing the ad (an approach that the researchers concede has its flaws—for one thing, there is no baseline of voters’ candidate preference prior to viewing the ad). One noteworthy finding so far has been that voters are hardly more favorably inclined toward positive ads than negative ones—they find positive ads no more believable, for one thing. Most striking of all though, for the purposes of understanding the state of the race, was the relative impact of the two campaigns’ ads. The researchers are trying to zero in on the impact of the ads among "pure independents," independent voters who, unlike so many purported independents, say they do not lean to one party or the other. Among this sample, Romney generally holds a roughly 15 point edge. This edge moves up slightly among voters who have viewed a positive Romney ad or a Romney ad attacking Obama, and it drops slightly among those who’ve seen a positive Obama ad. But this edge drops sharply, to only three points, among voters who have seen an Obama ad attacking Romney. And most effective of all, the researchers found, was Obama’s "America the Beautiful" ad. The unusually strong impact of Obama’s negative ads is not necessarily just a commentary on their quality, but rather a reflection of Romney’s relative lack of definition in the public eye. Since more voters have formed impressions of Obama, ads attacking him seem to have less impact than those attacking Romney. The goal of the Obama campaign, as laid out well by John Heilemann a few weeks ago, is plain: to make Romney an unacceptable alternative at a time when swing voters might be inclined to vote for the Republican out of frustration with the slow economic recovery under Obama.

"The [Obama strategists] have these kinds of data—this is what’s driving the campaign," said Vanderbilt political scientist John Geer. "It’s very much like ’04. The Bush people did not plan, at least if you believe [Bush strategist] Alex Castellanos, Alex says ‘we did not plan to run a super negative campaign. But we found out after a couple of months that we could not move President Bush’s numbers, but we could move Kerry’s.’ So they turned from what they thought would be mix of negative and positive to a very negative campaign because that was the only thing that mattered."

Other YouGov polling results presented by George Washington University political scientist John Sides (and first discussed in a cover story I did back in January) gave further context for why the Obama attacks on Bain Capital and Romney’s offshore bank accounts and investments may be effective. Voters, Sides noted, are far more likely to judge Romney a very wealthy man than they are to say that of Obama, and they also judge him less committed to helping the middle class than Obama. And there is a strong correlation between voters judging Romney a very wealthy man and their assessment of his interest in the middle class. This could help explain what Sides sees as a very crucial sliver of the electorate—the six percent of Obama voters who disapprove of his performance but who think that he "care more" about them.

But the political scientists also presented plenty of numbers that should give Obama pause. UCLA’s Lynn Vavreck, for one, presented data showing just how dour voters’ assessments of the economy are—and showing that the correlation between voters’ assessment of the economy and their presidential preference has been steadily rising over the past year, an ominous trend for Obama. "This is going to be a close election," Vavreck said.

follow me on Twitter @AlecMacGillis