The Blogging Caesar says, "Not so fast!"  Why?  As I scoured the internet recently searching for polls to include in my projections, I found a remarkable pattern among polls taken in hotly-contested congressional districts currently held by the GOP.  These polls reveal how a sizable Democratic advantage in generic polls may not translate to a sizable gain in House seats.  Before you scream "GOP spin!" let me explain.

Several polling outfits have been asking two different questions when conducting their polls.  The first is the named head-to-head question such as, "In November, do you plan to vote for Republican [Candidate's name] or Democrat [Candidate's name]?  The second is the generic "Will you vote for the Republican or the Democrat in November?"  The responses to these two questions have been amazingly disparate.

In order to illustrate my findings, I've configured the following table.  This first column denotes the race in question.  The second column contains the result of the head-to-head poll in that district, and the third contains the result of the generic vote question.  In each case, a positive number indicates a GOP advantage.  As you can see, the discrepancy between the two results is stark.

NOTE:  I cannot reveal the sources for these polls because I found them behind a subscriber wall.  I will say this about them, however.  They were conducted by three different polling firms, and they were all taken sometime in the last 3 months or so.

This eye-opening trend leads to the following conclusion:  Generic polls may uncover a sense of dissatisfaction with the majority party in Washington, but they do not seem, as head-to-head polls show, to indicate a willingness to kick the bums out.  It's as though the electorate has grown weary of GOP control, but doesn't see the Democratic candidate in these closely-held GOP districts as a viable alternative. 

Going forward, I'm convinced the GOP may be able to retain their majorities in both chambers of Congress, even if they continue to lag the Democrats significantly in generic congressional polls.  Until head-to-head polls in the close districts begin to show comparable results, I'm not too worried about the generics.

Update:  I incorrectly published the Arizona poll as AZ 8.  It should have been AZ 5.