Wow.

Democracy Corps has a new poll out:

It is very hard to look at the most recent Democracy Corps survey in the 50 most competitive Republican-held districts finished last night and not conclude that we are headed toward a 1994 election – with the Democratic majority approaching that of the ‘Gingrich Congress’. The named Democratic vote for Congress has moved up from a 3-point lead to 7-point margin since Sunday, with the named Democrat for the first time moving over 50 percent (51 to 44 percent). For the first time, the Democratic candidate is ahead on average in the bottom tier of least competitive races. The generic congressional ballot has moved up to 11 points – up 3 points from Sunday and another 3 points from the week before.

The gap in interest in the election (those ranking their interest as “10” on 1-to-10 point scale) between Democratic and Republican voters has grown from 7 to 14 this week.

The trend in the final week on most factors, as we will elaborate below, favors the Democrats, including independents’ support, enthusiasm, and handling key issues, including taxes, and above all, the Iraq war. The 2006 election is rapidly moving toward being a referendum on Iraq.

We do want to underscore in this last survey conducted by Greenberg Quinlan Rosner for Democracy Corps some reasons why you should believe this is real:

This survey, unlike any other public survey, asks the congressional vote using the actual names of each candidate, meaning this survey fully reflects any advantages for incumbency. We also ask the generic ballot for a read of overall partisan sentiment, but named vote is more likely to tell us what happens.

This survey is based on a 1,200 sample of likely voters conducted over three nights in the real battleground districts and will likely provide a more reliable read on the state of the race in the battleground.

This survey was not conducted on Halloween, which is a notoriously inaccurate night for polling.

The survey asks job approval and the thermometer rating for the incumbent by name, which produces a very disturbing result for the Republicans. Both measures are net negative, despite the power of incumbency. Incumbent approval is only 43 percent, with more disapproving, and indeed only at 44 percent in the tier 3 races.

While the survey shows a shift to the Democrats in the congressional ballot (named and generic) this week, the respondents have not shifted a point on party affiliation or recall of their 2004 vote. That should emphasize that this is a real, rather than a too-Democratic, sample.

These voters identify with the Republicans by 2 points, but in the generic, want to vote Democratic for Congress by 11 points.

We conducted 400 interviews last night after two news cycles of Kerry stories. While the story clearly broke through – over 20 percent of respondents in an open-ended question on what is happening in the campaign mentioned him – the story has not helped Republicans, and the calling on the last night was no different than the results from the first two nights. It is the highest recall for the partisans, those identifying as Democrats or Republicans, yet independents show no interest in it.

The shifts in this final week are almost entirely among independents who are giving the Democratic candidates landslide margins.