One of the polls that I track is Rasmussen Report’s weekly generic congressional ballot. My log of poll results goes back more than six years, to January 2009. The following graph shows the results and compares them with the popular-vote margins in the House elections of 2010, 2012, and 2014:
Rasmussen’s poll was accurate in 2010, when the nationwide tally of votes for House candidates favored the GOP by almost 7 percentage points. The poll was accurate again in 2012, when the GOP came up short by more than 1 percentage point. But the story was different in 2014, when the poll undershot the GOP’s victory margin of 6 percentage points.
Is it possible that the poll sample has become biased toward Democrats? Perhaps. It’s also possible that Barack Obama’s lingering unpopularity gave the GOP an extra boost in 2014. (See this post.)
In any event, if the generic congressional ballot means anything, it means that the GOP is still on the upswing. It’s halting and slow, to be sure, but it’s an upswing. The question is whether Boehner, McConnell, and company can capitalize on their party’s relatively good standing with voters.