In its poll, the Democracy Corps surveyed voters in 54 House districts, equally divided into tiers of 27 seats that the group rated as the top and secondary Democratic opportunities. In 2008, Obama carried 25 of the 27 seats the group rated as the top Democratic targets and 10 of the 27 it placed in the second tier. The Democracy Corps survey (conducted from July 21-26) found Obama leading Romney by 52 percent to 44 percent in the top tier of Democratic targets and trailing by 48 percent to 44 percent in the second tier. Overall in all 54 districts, Obama led Romney 48 percent to 46 percent, which nearly equals his four percentage point overall 2008 margin in these seats, the group reports. This is the first Battleground poll conducted entirely within the new district lines; the last one conducted in April questioned respondents within the 2010 district lines. To ensure as comparable a sample as possible, the pollsters adjusted their April results to reflect only districts that remained mostly untouched under the new lines. Obama's improving position looms so large because House elections are now so heavily influenced by attitudes toward the president. In 2004, exit polls showed that 88 percent of voters who backed George W. Bush also voted Republican in House elections and 89 percent who voted for John Kerry voted Democratic. In 2008, 90 percent of Obama voters backed House Democrats and 86 percent of John McCain voters also pulled the lever for House Republicans. Even without Obama on the ballot, the relationship held in 2010: according to the exit polls, 85 per cent of voters who approved of Obama's performance voted Democratic for the House, while 84 percent of disapprovers voted Republican. In these competitive districts, the Democracy Corps poll finds the races swept along by those same currents: it found that 89 percent of Obama voters intend to vote Democratic for the House, while 93 percent of Romney voters say they plan to vote Republican. Given the intensity of that correlation, Greenberg says it wouldn't be surprising for Democrats to win two-thirds of the 27 top-tier targets in the survey, many of them affluent suburban seats that were carried by Obama in 2008. That would be comparable to the GOP performance in 2010, when it won fully three-fourths of the 48 House seats held by Democrats in districts that voted for McCain two years earlier. (Another 22 of the GOP pick-ups that year came in districts where Obama won narrowly in 2008, with 55 percent of the vote or less.) But those gains alone wouldn't be enough to provide Democrats the 25 seats they need to regain a House majority -- especially since Republicans appear positioned to capture at least some additional seats still held by Democrats in districts that voted for McCain. After redistricting, Democrats hold 14 such seats, and several of them -- such as four Democratic-held seats in North Carolina and one each in Oklahoma and Arkansas -- are clearly at risk this time. That means to regain control Democrats will almost certainly have to win some of the "Tier Two" targets -- whether won by Obama or McCain in 2008. And there, as the Democracy Corps poll shows, the terrain remains rockier for the party. In the survey, not only does Romney still lead Obama in those 27 seats, but Republicans maintained a nine percentage point overall lead on the generic ballot. Greenberg says the key to overcoming that advantage will be to intensify the focus on the Ryan budget, which converted Medicare into a premium support, or voucher, system while maintaining large tax cuts. "You cannot look at this data without realizing they have increased vulnerability with the Ryan budget," he argues. Lindsay argues that Greenberg is overestimating the capacity of the Medicare issue to neutralize the other GOP advantages in those districts. "How many times have we heard them say this?" he said. "Of course when it's phrased [their] way it moves numbers, but it also moves numbers when we say a Democrat who voted for Obamacare voted for $500 billion in Medicare cuts. I'd just argue there's no real [Medicare] advantage on either side." As these cross-currents intersect, the Democracy Corps poll raises the prospect of a House that is even more polarized after the election -- with virtually all Democrats representing seats that also preferred Obama and nearly all Republicans representing constituencies that rejected him for Romney. Regardless of who wins the White House, that's a formula for a deepening of the quasi-parliamentary patterns that have produced the highest level of congressional party-line voting since the late 19th century -- and unstinting partisan conflict on almost all major issues.