The title of this piece might seem a bit counterintuitive given the presumed certainty of Republican gains on November 2, but within that context, there really is a surprising amount of uncertainty about which party is likely to get the late breaks in this cycle.
On the one hand, state polling is showing some good signs for Democrats in Senate and some gubernatorial races. Two left-for-dead candidates, Joe Sestak of Pennsylvania and Russ Feingold of Wisconsin, have rebounded into highly competitive positions, according to some polls. Joe Manchin of WV seems to have recovered from a near-fatal swoon. Poll numbers for Richard Blumenthal of Connecticut have stabilized, as they have (at a lower level) for Patty Murray of Washington and Barbara Boxer of California. At least one poll shows Robin Carnahan of Missouri with a mini-surge, and Michael Bennet of Colorado seems to have drawn even with Ken Buck. The brief period of hysteria about a possible Tea Party takeover of New York politics has ended in derision. And at the moment, Democrats are optimistic about winning at least one southern governorship, in Florida, and believe they have an outside shot in Georgia and (surprise, surprise) South Carolina as well (polls are showing Nikki Haley losing support and making the race competitive).
But at the same time, certain meta-indicators are ominous for Democrats. Gallup’s last two generic congressional ballot tracking polls have shown Republicans with double-digit leads among likely voters, an unprecedented phenomenon. Worse yet, in a low-turnout scenario, Gallup has Republicans up by 17 percent, which if accurate would produce House gains well above what most analysts have been talking about. And Gallup’s not alone: another highly respected research firm, Pew, put out its own generic ballot poll this week giving Republicans a ten-point advantage among likely voters.
So how can we explain the macro-micro disconnect in polling at this moment? It’s possible that Gallup and Pew just have it wrong (Alan Abramowitz of Emory University has charged Gallup with making crucial errors), and that other generic polls will soon demonstrate that those results are outliers. Another common theory is that statewide races operate according to different dynamics than overall partisan preferences, and that while Republicans may make big House gains, that doesn’t necessarily translate into victory in close statewide races.
At RealClearPolitics today, Sean Trende suggests it’s the state polls that may be off, thanks to inadequate likely voter screens that are modeling the electorate’s partisan composition too favorably to Democrats. Using a partisan composition model based on the two 2009 gubernatorial contests, Trende hypothesizes that Republicans statewide candidates may on average perform better than their polling by a 3-4 percent margin, which would, of course, throw many close races to the GOP.
Complicating all this analysis of public opinion research, of course, is the fact that the two parties’ ground games are just now really kicking in, which could change turnout patterns, along with the phenomenon of very heavy early voting. On this latter front, the preliminary data indicates that Democrats seem to be doing a relatively good job of early voting mobilization, but don’t have the sort of advantage they enjoyed in 2008, and may not have an advantage at all in certain key states (e.g., Colorado, Nevada and Florida).
Then you get into some really hazy phenomena that may affect particular races. The most discussed is California’s Proposition 19, which would legalize small-scale cultivation and use of marijuana. There is a persistent belief among California Democrats that Prop 19 will turn out younger voters (and perhaps African-Americans and Latinos) at higher levels than in other states, giving Democrats a crucial boost in close contests.
But overall, the varying indicators of late trends (unless unanimity suddenly emerges between now and November 2) are providing some real mystery and drama in this bitter cycle, and plenty of questions to mull over in the post-election rumination period that will ensue.
Photo credit: bjornmeansbear
It’s another Tuesday, and believe it or not, there are no primaries today! In fact, the next batch is not until September 14, when seven states plus the District of Columbia hold elections. This last weekend, however, voters in Louisiana and West Virginia went to the polls, with the latter limited to a special primary election for the late Robert Byrd’s Senate seat.
Today’s major primary is in Georgia, and I covered the Peach State contests pretty thoroughly last week (for more detail, see this 
