I especially like the >50% part.

This is Bobby Bright.  There might be people vying for the Republican nomination right now, but effectively Bright is also the Republican candidate in this race.

He's one guy I wouldn't mind seeing be unemployed.

So far polls show us trailing in MI-07, OH-01, NH-01, MD-01, IN-09, NM-02, and we will probably lose open seats like AR-02, KS-03, TN-06, LA-03, and PA-07(thanks Sestak).  That is 11 seats lost and if we beat Cao in LA-02 and pick up the open seats DE-AL and IL-10, we can hold our losses to eight seats.  However, there will probably be a number of upsets. 

Charlie Cook gets paid for being right or wrong, and still has plenty of time to backtrack off of this latest hyperbole to maintain his relationship. Trust the polls, not the pundits.

guesses at the 33 most vulnerable Democratic-held House seats:

http://www.swingstateproject.com/showDiary.do?diaryId=6452

I would guess at most four or five Republican-held House seats are vulnerable, so to win back the House the GOP would have to sweep this whole list plus pick up 10 or 12 House seats that aren't on the radar yet.

If the economy doesn't improve, though, that is very possible.

Jonathan, I think what this poll suggests is that Democrats in tough districts who have opposed the Democratic Congressional leadership and the President on just about every important matter, have a decent chance of surviving. 

But I think that you might agree that very few of the 15 other House Democratic incumbents that we (www.cookpolitical.com) have rated as Toss Ups, few of the 26 incumbents we have listed as Lean Democrat (our two competitive columns for Democratic incumbents) or for that matter, nto that many of the 40 Democratic incumbents we have listed as Likely Democrat, our "watch"  list, have opposed the party, that is, put as much distance between themselves and their party, as much as Bright has.

I'm not sure you really want to make that point, do you?

I should have noted that this past week National Journal released their Congressional voting ratings, a statistical analysis of 191 House votes last year, finding that Rep. Bright was the most conservative Democrat in the House.  If you want to hold Bright up as an example of a Democrat in a tough district that seems to be doing pretty well in his fight for survival, I will agree with you.  Having read some of your writings before though, I'm  not sure that's the point you want to make.

I also might point toward the midterm Congressional election model that Emory University's esteemed political scientist Alan Abramowitz released last week on Larry Sabato's Crystal Ball site, projecting Democratic losses of 37 or 38 seats, an eyelash short of the majority turning over and not inconsistant with what we are saying. 

And while you may not like what we are writing these days, since we were the first to write (http://cookpolitical.com/node/2345) on  August 5, 2006 that it looked like the House was going to turn Democratic and that the Senate was 50-50, you might give us a little credit for maybe not being crazy.

The bottom line is that you are leaning on a very thin reed, using this one race to suggest that the Democratic majority is not in grave danger as I have been suggesting for months. 

My job, having started the Cook Political Report in 1984, is to call them as I see them.  We saw a big wave coming in 1994 but underestimated it then.  In 2006 we saw one and nailed it.  We saw signs of problems and began writing and talking about it last summer and see little sign that we are wrong.  If more Democrats had the cover that Bright had, maybe we would be.

Seems to me a large part of the alarm about the coming Dem losses can be traced directly back to command of narrative, Charlie.  For whatever reason, the media always seem to feel that every turn if events is "bad for Democrats".  The coverage shows it, the stories, the punditry, the analyses, the conclusions show it.  The GOP has shown a remarkable ability to put their thumbs on the scales in Editorial suites across the media spectrum.

It is not so much about the candidates, the races, or the issues, but about the media only being focused on increasing their market share by playing to the most emotionally hot headlines.

I guess " Dems in trouble!!"  sells papers and garners pageviews today, and seems to be the most likely direction for editors to go with that day's content.

Republicans will tell you they screamed much the same thing when Bush was being dragged down to 20% approval in 2008, and it isn't any easier for Dems like me to take it today when the shoe's on the other foot.

Looking forward, I think we have to wage a better, more focused, more organized war of words.

We have to give the editors some quality stories to write about that will help them sell more papers and gather more eyes, and will give us in return better stories about better Democrats that will help us swing races our way instead of giving them away to the opposition.

Yes, we have been writing "Dems in trouble"  for about eight months now, but it isn't much different from when we were writing that Republicans were in trouble during the 2006 and 2008 cycles.  And Republicans were in fact in trouble.  Our job is to watch races individually and look for trends.  If the partisans for the side on the short end of the trends don't like it, they typically attack the messenger.  you can be sure that Republicans weren't happy with what we were writing in the months leading up to the 2006 election, but we were right.

To the previous question about Iowa 3, Leonard Boswell's rating change from Likely to Lean Democrat, our ratings are not purely based on public polls (in the House there aren't that many), but also on unreleased polling, on interviewing experienced political operatives and reporters in each state as well as national party strategists.  Our "Likely" Democratic and Republican categories are basically "watch" lists, contests that could become competitive, but aren't yet.  the "Leans" are those that have become competitive but were there still is a clear leader.  We now see Iowa 3 as a competitive race based on all of our research and reporting.  But the short answer to your question is that we do not rely exclusively on published polls.

I tend to agree with most of Cook's race ratings (aside from his policy of never rating incumbents below "toss-up"), but I also feel that he is affecting the narrative by making statements like this in absence of much actual polling data. This is similar to what Rasmussen is doing, flooding the conversation with data that actually alters the narrative in a specific direction. They build a GOP advantage into their polling, publish the results, and suddenly there's a huge GOP advantage! Everyone flips out, Dem incumbents start retiring, strong challengers back away, and fundraising slows, while the GOP gains strong recruits everywhere. It's no secret where Rasmussen's loyalties lie, so this clearly is the intended result.  It's not about polling or predicting; it's a deliberate, partisan effort to affect the next elections, and it's incredibly effective. Rasmussen alone is probably responsible for strong recruits and retirements in several states; that's far more than the NRSC could have done alone.

If I recall correctly, few pundits really thought the Dems would take the Senate in 2006 or achieve 60 seats in 2008 (although no one was counting Specter at the time). Everyone seemed far more cautious about broad statements back then. This season, the conventional wisdom shaped the election results before the current Congress was even sworn in. A lot of assumptions were made right off the bat, like "the party that controls the White House always loses seats in the first mid-term," and "anyone in marginal districts who votes for the Obama/Democratic agenda will lose in November," and "you have to vote like a Republican to get re-elected in a Republican district." None of these things were true on January 20, 2009. They didn't become true until the Democrats utterly failed at governance and allowed the GOP to control the message on every issue.

My main point is that Charlie Cook, Rasmussen, and many other pundits, pollsters, or what-have-you, have a strong impact on shaping the narrative of the election. Regular people pay absolutely no attention to them, but politicians, donors, and other politically engaged people do. When they say something is true, that tends to make it become true, even if it was completely false before. That's a rather dangerous scenario when their messages are based on partisan exaggerations (Rasmussen, Rothenberg), pure conventional wisdom (CQ Politics), or vague data that can't be quantified, measured, vetted, or even defined clearly, referred to here as "research and reporting" (Cook). I realize there needs to be some method for rating races based on data besides published polls, since there often are none for many House races, but I feel like this is a "gut feeling" more than a scientific assessment. More transparency couldn't hurt.

When bpfish says, "If I recall correctly, few pundits really thought the Dems would take the Senate in 2006..." all I can suggest to Mr/Ms Fish is to look at the August 5, 2006 I provided above.  Saying three months before the election that Democratic chances of picking up control of the Senate was 50-50, they needed six seats, they got six seats, by a margin, if I recall, of less than 30,000 votes nationwide.  Doesn't get much better than that.  I just don't get where this all comes from.  We've been dead on the last two elections, yet there is some revisionist history that nobody got it right.  Take a look. Our ratings and years of my back columns are at cookpolitical.com.