Yesterday, Daniel wrote a diary,"Did the GOP Really Pull Even With The Democrats in the Congressional Ballot?", which was especially critical of the internal demographics of NPR and Rasmussen, thereby questioning whether the Republicans had actually closed the generic ballot gap to virtually nothing. In comments, I referred to the DKos/R2000 tracking poll as counter-evidence to the supposed trend. Neither Daniel nor I are suggesting that the polls should be ignored completely. One should always be quite reluctant to throw out a poll entirely. But putting polls into the "I don't know about that" category is a different matter, and I'd argue that that seems to be what's warranted here.
To make this point, here's the summary data of DKos/R2000 tracking poll for Congressional Reps and Dems over the past two months (more detailed tables on the flip):
As can be seen, there's only a very slight trend on the GOP side, while the Dems show a strong gain outside the South in February, which faded somewhat in March, still leaving them stronger than they were at the beginning of February. While this is not the same as a generic ballot poll, it's obviously a related measure, and these results are not compatible with the generic ballot narrowing to a dead heat.
On March 17, NPR published a poll under the innocuous headline "More Voters Think U.S. Is On Right Track." That poll had a rather surprising result on the question of generic Congressional ballot preference, with Democrats and Republicans tied at 42, and 15% undecided. Whoa. Is that true?
Well if you check the pollster trendlines, this seems entirely plausible:
Taking into account the districts with open seats and the current generic ballot advantage held by Democrats, we would expect to see around 21 Republican-held House seats flip to Democrats this year if conditions are similar to the elections of 1994-2006.
Below, an explanation, as well as details on party identification, approval, House retirements, and the generic ballot. First up, party favorability ratings:
