Gallup has held out quite a long time from imposing a likely voter screen on its generic ballot, but that will be coming to an end next week. The pollster suggests that this could result in a massive swing toward the GOP, perhaps even showing a double-digit GOP win:
Additionally, preliminary modeling of the likely electorate using Gallup’s traditional likely voter questions (more on this next week) suggests that if current patterns persist, Republicans could have a double-digit lead in the national House vote on Election Day, which would translate into Republicans gaining well above the number of seats necessary to control the House.
The reason? Gallup has basically been asking people only if they are registered to vote. If they are, Gallup proceeds to the ballot test questions. Next week, the likely voter screen will begin probing how frequently these respondents vote, and more importantly, how enthusiastic they are about voting. This enables Gallup to predict which registerd voters will turn out in the fall. Republicans are substantially more enthusiastic about voting than Democrats right now, so this will likely move the generic ballot substantially in their direct.
To put it differently, the RV polls in the current RCP Average show a 1.75 point Republican lead, while the likely voter polls currently show a 6.25 point Republican lead.
For a point of reference, in 1994 the GOP won the national vote by 7 points and held 230 seats on election night. In 2006 the Democrats won the national vote by 8 points and finished with 233 seats. And in 2008, the Democrats won the national vote by 10.5 points, and finished with 257 seats. 257 Republican seats would translate to a 78-seat pickup.