Crosstabs.org

Last Friday (10/6) Steele’s campaign released polling conducted in late September that shows him trailing by just 5 points – – 44% to 39%. Although I’d love to believe that Steele is within 5 points, it doesn’t square with an independent poll conducted earlier in October that showed Cardin’s margin at 15 points – – 54% to 39%

This November 7th in each of 435 districts, voters will elect a Congress critter. The WashPost asserts that “Republican campaign officials” tell them that they expect to lose between seven and thirty seats to the Dems on that day. These 7 to 30 races, the Republican campaign officials tell the WashPost, might to Dem because of “nonstop controversy over President Bush’s war policy and House Speaker J. Dennis Hastert’s handling of the page scandal.” (The high number, 30, came from the lips of Tom Davis. His motive?) What we have there, really, is the MSM boasting about the power of the MSM. It is the MSM, after all, which have generated and promulgated this “nonstop controversy.”

Reuters sees the end of the Republican majorities: War, sex scandal sinking Republicans in polls:

The latest polls give Democratic candidates a growing edge of between 13 and 23 percentage points over Republicans on the November ballot, with Bush’s approval ratings dropping back into the 30s after a slight bump into the low 40s in September.

President Bush is not seeking reelection this November. Some swaggering, smirking media caricature named Republican is not facing some svelte do-gooder named Democrat for control of Congress. The admonition for local Republican candidates to concentrate on local issues and leave the national stuff to the national party is as sound as is the national party.

You tell me.

(Read More…)

This is as close to pollster hell as it gets.

About a week ago we were starting to see evidence in national polling that Republicans were turning the political tide that had been working against them the past few months. As two observers said:

Dan Balz, The Washington Post 9/24/06
“After months of unrelenting bad news, President Bush and his Republican allies have begun to change the mood, if not the overall trajectory, of a midterm election campaign that has tilted against them for a year.”

Charlie Cook, National Journal 9/26/06
“If Republicans can replicate the environment of the past six weeks, their chances of holding onto their House majority are pretty good, and they will almost certainly retain their Senate majority.”

Majority Watch has done some polling in FL-16 in the aftermath of the Foley scandal. They did something interesting– they did two 1000+ voter samples, asking the horserace question in different ways. They explain here. Basically, in one sample they simply gave the candidates, Foley and Mahoney. The results:

FL-06, Majority Watch, Mahoney (D) 50%, Foley (R-Incumbent but oh-so-outta-there) 43%. Poll conducted October 1 amont 1001 likely voters.

And among those certain to vote, it expands to 52%-42%.

SurveyUSA has a new poll that has Sen. Allen with a five point lead over Jim Webb. The entire poll was conducted after the latest bruhaha. But, in what is sure to be the headline grabber (hence the headline of “Virginia Senate race in flux”), SurveyUSA wrote the following:

In an election for the United States Senate in Virginia today, 9/27/06, incumbent Republican George Allen maintains a slight 49% to 44% advantage over Democrat challenger James Webb, according to a SurveyUSA poll. Since an identical SurveyUSA poll 2 weeks ago, Allen is up 1 point and Webb is down 1 point. BUT: day-to-day data shows that the race is volatile. On Sunday 9/24, after Allen had been accused of using racial slurs in college, he led by 7 in SurveyUSA Sunday-only data . On Monday 9/25, after Allen strongly denied the accusations, he led by 11 in SurveyUSA Monday-only data. On Tuesday 9/26, after more people corroborated the accusations, Allen trailed Webb by 3 points, in Tuesday-only data. The 5-point Allen advantage shown here, when the 3 days of data are combined and averaged, cannot be considered stable.

We often caution people to not read too much into any one survey, and that is advice that still applies. Right here, though, the SurveyUSA analysis ventures into the land of reading too much into a small portion of a single survey. If about 1/3 of the sample was taken each day, then the margin of error for that one day sample is +/- 6.7% per candidate.

On the heels of last night’s debate between Republican Rick O’Donnell and Democrat Ed Perlmutter in Colorado’s 7th Congressional district, KUSA-TV is out with a new SurveyUSA poll showing Perlmutter with a commanding 17-point lead in the race, 54-37. That’s a huge jump from the last SUSA poll in August showing the race a 45-45 tie. All caveats about putting too much stock in any given poll still apply, so be on the lookout for further polling to see if this race has really gotten away from O’Donnell.

All caveats, indeed! First, this poll was released on the heels of the debate, but it was taken before it. But that’s not what really caught my eye here.

I looked at the raw numbers in both the August iteration that showed the race a toss-up and the new one showing a Democratic rout. In both, Perlmutter had an overwwhelming amount of support from Democrats, losing 4 points to O’Donnell in both. Then I noticed that Perlmutter had a 25 point advantage with self-described independents. Hm, not much change there, either. The two polls taken together do show a slip in Republican support for O’Donnell, dropping from 85%-10% to 76%-14%; that fits the meme I have been pushing which is that it is all about the GOP getting self-identified Republicans to come home to the party by election day. However, that amount of change in Republican support cannot explain the diffference between the two surveys’ topline results.

The earlier survey consisted of 44% Republicans and 33% Democrats. The newer survey consists of 38% of each. Given the nearly monolithic support Perlmutter is getting from self-identified Democrats, the change in partisan composition in the surveys would account for most of the movement shown between the two surveys.

Why this occurred is anyone’s guess. Was there widespread changes in the self-identification of people in the district? Possible, but not likely. Were normally loyal GOPers reluctant to participate in the survey this time around? More possible– and then the question is if it was a one-time phenomenon due to other things happening (such as the Broncos playing) when the poll was conducted, or does it mean that Republicans have lost interest in the race? Were normally loyal GOPers overly represented in the August sample? Possible, but not likely; the district has more Republicans than Democrats.

Obviously, it is much better to have good results for your candidate than bad results. That said, this looks like one of those cases where surprising results were due to the sample having a mix that is less representative of the whole than normal.

Democracy Corps, the Democratic group founded by James Carville, Bob Schrum, and Stanley Greenberg with polling done by Greenberg Quinlan Rosner, has a new “Report on Public Polling” out. In it, they examine the recent trend of improvement for the Republicans in public opinion surveys. “An objective analysis of averages derived from all of the polling conducted and released since Labor Day,” they write, “as well as trends on a number of key measures, indicate there is undeniable truth to the notion of Republican gains.”

So I’m looking over some political news and I come across this piece talking about a new poll in IN 8 showing GOP incumbent John Hostettler trailing his DEM challenger, Brad Ellesworth, by 15 points. Really this isn’t too shocking as Hostettler is a perennial incumbent on election watch lists every cycle. However, the survey probably isn’t worth the paper it’s printed on.

New from Rasmussen ($) today, Jon Tester (D) leads incumbent Conrad Burns, 50-43. This is a very slight change from the last poll, which had Tester up 52-43. The “gravity” of Montana will probably pull this race towards Burns as the election draws near, but Tester being above 50 in two consecutive polls does not bode well for this race.

Also, Maria Cantwell’s bounce in the wake of the McGavick revelations is gone – McGavick is back within 6, 48-42. Cantwell is facing her own problems with a dodgy loan made to a lobbyist, and continues to face high unfavorables.

Also just released in the PA Senate race, a Keystone Poll (warning: .pdf) has the race 45-38 in favor of Casey. This represents a very small shift from the August Keystone poll, which had it 44-39 in favor of Casey, but the internals are a little more troubling. In August, the polling sample tilted very slightly Democrat, 45-43. In this poll, the sample tilted very slightly Republican, 46-43. Nonetheless, Santorum lost two points in the overall result. Combined with the Rasmussen poll released today, it seems clear that Santorum is not getting the post-debate bounce that he hoped for, and he continues to languish in the high 30s.