US Senate race polls: Gains in NJ & RI, setback in OH

Senator Robert Menendez (D-NJ) has widened his lead over challenger Tom Kean Jr. (R) according to a June 26 Rasmussen poll. Menendez now leads Kean 46% to 40%. A June 22 Rutgers/Eagleton poll gave Menendez a 42% to 38% advantage.

Challenger Sheldon Whitehouse (D) now leads incumbent Senator Lincoln Chafee (R-RI) 38% to 37% according to a June 26 Brown University poll. A June 21 Rhode Island College poll gave Chafee a 43% to 40% lead. If Chafee loses the close primary race against Stephen Laffey, Whitehouse's prospects look even better.

Senator Mike Dewine (R-OH) now leads challenger Sherrod Brown (D) 46% to 39% according to a June 27 Rasmussen poll. A June 21 WSJ/Zogby poll had given Brown a 46.7% to 34% edge.

Overall, Democrats now stand to pick up 4 of the minimum of 6 seats needed to regain control of the Senate: Missouri, Montana, Pennsylvania, and Rhode Island. Close races in Arizona, Ohio, Tennessee, and Virginia give Democrats reason to hope that this goal is attainable.

yet again.

Did you guys know that Rasmussen is based out of NJ? That's my state, and I have petitioned the (Democratically controlled) state legislature to begin hearings on how it conducts it polls, but so far they haven't. NJians, you may want to start paying attention to this...

.

Hate to support CTs, but... :tinfoilhat:

Just because you don't like the results of a poll doesn't mean that its wrong or corrupt. Rasmussen has a different polling method than other pollsters; in fact, they all vary. In some races, Rasmussen is the most accurate. In other races, other pollsters and different methods are more accurate.

.

I would not trust that one.

That Ohio poll is bogus .... but you can be sure that DeWine will wine here.
The corruption level is mind blowing.

I just take the very latest poll into account, without any regard to their reputation. It does seem to me that Rasmussen is right-leaning in general, but you might want to consider that they poll likely voters, which tends to give Republicans an edge. All polls do not limit their surveys to likely voters. But like I said, I'm just tracking the latest poll in each race.

If you want to compare all polls that have been conducted in any of these races, here is a good resource: http://en.wikipedia.org/wiki/Category:United_States_Senate_elections_by_state The information there is being accurately updated in a timely manner.

The polls are crazy - swinging from one extreme to another on this race, while being pretty consistent in showing Strickland over Blackwell in double digits. Something doesn't make sense. The Zogby poll is also a poll of "likely voters" so that does not explain the discrepancy.

You are very right about the corruption level. If any Republicans win statewide races this fall, I guarantee you it is because of election fraud. We are all so disgusted with what is going on. Four more Republicans indicted today in the Noe-BWC-Bush scandal. It's appalling to everyone - Rs included.

Also, everyone should remember that just as in the U.S. as a whole, there are more registered Dems in Ohio than registered Repugs. Nothing is adding up.

That is the story, all those big homes outside the suburbs put bush over the top.
That is the cover story "they" are sticking with along with the value voters too.

(but it was new voters who put bush over the top and Ohio registered 350,000
new Kerry voters to 30,000 new bush voters ........ you are right nothing is adding up.

BTW Welcome to DU!

How could the Ohio poll swing 21 points in 6 days, between Zogby/WSJ & Rasmussen? There is an obvious difference between the two polls' results. Zogby has released 3 polls, and all have predicted Brown would win. Rasmussen has conducted 7 polls since December last year. Only one of the Rasmussen polls picked Brown as the winner. Survey USA and the University of Cincinnatti have conducted one poll apiece. Survey USA resembles Zogby conclusions, while the U of C poll seems to suggest the Rasmussen polls are more accurate.

A comparison of respective polls' 2004 predictions with actual reported election results shows Survey USA to be the most accurate, if memory serves. Please don't take this as gospel; I'm not sure and don't feel like looking it up right now.

But then the question: Of the polls that most accurately predicted the 2004 election results, didn't they actually exhibit a Republican bias since the election was stolen? My answer to that is yes. With reluctance because it is so profound, I have come to the conclusion that the 2004 election was indeed stolen by Republicans. It would be very interesting to see all these polls' 2004 predictions lined up with the official vote tallies, and also with exit poll results that were reported before the corporate news media 'adjusted' them to conform with the official election results.

But I'm not going to do that right now because it would be time-consuming. That might be a good project for you or someone else here, or maybe I'll decide to tackle it someday. I hope you can understand why I've decided to track these Senate races by using the most recent polls conducted by whomever, without any other regard concerning the source. I think it's interesting and a little fun that way.

Their state polls in candidate vs. candidate races seem to be very reliable. Ballot inititative polls, not as reliable.

Numbers are all over the place right now. You can find Brown well ahead or similar for DeWine. Same thing for Talent and McCaskill. I'll conclude those races are very close to dead even, which makes sense. No way we are burying senatorial incumbents in red states, incumbents without Santorum-like idiocy or abysmal approval ratings.

Thanks, Awsi. But for what it's worth, I'll continue to hope we have a chance to pick up 6 or more Senate seats this November. I might get disappointed but I gotta hope.

House races are harder to get your arms around but I believe we have an even better chance there.

Lasher

even competitive.

That they actually control the Senate now shows that the American people are totally out of their fucking minds.

which is basically the strongest issue for the repukes.
If dems come up with taxcuts in their platform, it would be
a landslide in favor of dems.

what it is. A once proud and good state seems to have degenerated into a political cesspool.

I've counted alot of Brown bumpers stickers here near the Ford plant in Elryia. I'm only here for a day but I was impressed. I think the UAW is doing a big push in this zone. The polls are swinging to wildly between Dewine and Brown very, very strange

showing Brown continuing to pull ahead. He has a double digit lead now.

Rasmussen also had some weird results in their last poll for Strickland, showing he was down 3 points or more. Baloney. There's no way Blackwell is gaining in popularity in Ohio.

When calculating "likely" voters, Rasmussen's sample must be excluding all those precincts Blackwell plans to purge from the voter rolls.

the margin of error? I remember seeing that here a few days ago.

Rasmussen has been off on the last several polls in this race, not sure what they're trying to do.

When the exit polls again point to massive corruption I won't be surprised.

It has been consistently way out of synch w/ other polls on this race. Their polls vary by a huge amount from others done during the same time period. Not sure what they're up to or what they're doing wrong, but they're way off from other polls.

Zogby and Survey USA have been consistent, showing Brown continuing to increase his margin, now in the double digits. He's doing fine.

Hell, Rasmussen isn't even looked upon as one of the major polling services..

Scott Rassmussen is a wingnut who writes for a conservative fishwrap when he's not cranking out his pathetic polls.

There are various ways to evaluate which polling firm had the best results, but Rasmusssen was near the very top of the list via every measure in '04, along with SurveyUSA and Mason Dixon. His numbers in battleground states were very close and there wasn't a partisan slant either way. That's comparing official vote count to pre-election poll.

I've found a good number of people who can objectively and rationally discuss matters like campaigns and polling numbers, but there are always those who refuse to let reality set in.

Not only for supposedly tilting toward Republicans in his polls, but the way he poses the presidential approval/disapproval question seems to lead to a higher approval number than typical.

Actually I haven't looked at the approval question in a long time but someone here pointed it out to me in '04 and the question did differ from how others phrased it.

There are a handful of DUers who subsribe to Rasmussen's pay service and post those poll numbers here shortly after release. I assume that will hold up this fall.

Which can't be found on his own website of course.

The guy has more ties to major wacko rightwing organizations than Ann Coulter does.

Someone posted a great review of polling agencies that was created by one of the major newspapers a while back, and Rasmussen (while you may have liked his 2004 polls) was not considered one of the top tier polling agencies, because they said he wasn't non-partisan.

I realize he gets it right on occassion ... they all do. But pull up some articles on the man himself SCOTT RASMUSSEN and read about some of the extremes he's gone to to skew polls for candidates he's been hired by.

If he ran a non-partisan polling site, he might make become one of the big boys like Gallup, etc..

But not Scott Rasmussen.. no way.

If that's enough to disqualify a polling agency, then there's really only a small handful you can look at, and I don't think that's sufficient to get the big picture.

Discredit and ignore Rasmussen all you want, but do so at your own peril. There's certainly no logical basis behind it.

You are not correct to say there is no logical reason to suspect that Rasmussen has a bias in favor of Republicans.

As Awsi Dooger mentioned in reply 38 of this thread, his Rasmussen's presidential approval rating is almost always higher than every other poll. I say almost, because I can remember only once or twice seeing him tied with one other poll for first place. Here is a link to Real Clear Politics, where you can view their composite presidential approval rating:

http://www.realclearpolitics.com/polls/

Once there, if you click on See all Poll Data you can review historical data.

Awsi supposed that this difference might be due to the way Rasmussen's question is posed to respondents, but his poll weighting is a more likely cause. As divulged only recently, prior to May of this year Rasmussen surveyed an equal number of Democrats and Republicans. It is quite obvious that this has cause a pro-Republican bias because there are simply more Democrats. Rasmussen now uses a weighting of 36.6% Democrat, 33.5% Republican, and 29.9% Unaffiliated.

http://www.mysterypollster.com/main/2006/05/rasmussen_adopt.html

Awsi in reply 42 that he considered Gallup's 2004 pre-election equal poll weighting to be appropriate, based on research by Pew and others. This is not an unreasonable point, but I disagree. One reason is that historical exit poll results indicated pro-Democrat weighting was appropriate, as he pointed out. He went on to say that 2004 pre-election pollsters guessed right when they use equal weighting (this includes Rasmussen). I don't think they did because I believe the 2004 election, and its exit polls, were fraudulent. But now I'm getting into a more highly debatable arena so I will leave it at that.

In reply 39 larissa raised questions concerning Rasmussen's track record, and claimed she knew of cases where he had intentionally skewed poll results in favor of clients. I don't have time to research that myself today, but I'm not going to dismiss her point until I do.

These are reasons why I think it's reasonable to suggest that Rasmussen might have a pro-Republcan bias - or if you prefer to think of all polls as biased to one degree or another, more highly biased than most others. I agree that we shouldn't ignore Rasmussen but you're wrong to say that those who discredit his findings have no logical basis to do so. There are good points on both sides of this argument.

I check that site almost daily but completely missed that link in May. I had no idea Rasmussen had become more specific about his weighting techniques.

That +3.1 party ID edge for Democrats that Rasmussen is using now seems accurate to me. I have Excel election spreadsheets and I've been using +3.3 for a national base, then adjusted for the individual state.

The problem right now, as in 2004, is that number is hardly stagnant or obvious. Prior to 9/11 you could just leave it alone. Post 9/11, many reports came out indicating a switch toward the GOP, which made sense due to national security as a priority. As that Mystery Pollster link mentions, there was also a spike in party ID toward the GOP after the '04 Republican convention. PEW itself had two different party ID findings in '04, one indicating a continued drift toward the GOP while the other said Democrats had regained the edge, almost to pre-9/11 levels.

It was hardly surprising that the polls in 2004 varied greatly and were in high dispute, largely due to that party ID weighting. The pollsters giving the highest pro-Republican numbers, like Gallup, were sensing and reporting an even greater party ID shift than PEW or others had indicated. They believed the plurality of November voters would favor Republicans in party ID so their polls matched. The pollsters sticking with tradition applied the +3 to +4 Democratic edge. On DU we thought that was correct and we loved those polls, but on average they turned out to be the most innacurate. Obviously I'm accepting the official vote count in this argument, not the exit polls. That can be debated elsewhere, as Lasher implied.

Again, I think the pollsters applying even party ID in '04 got it correct mostly due to guessing right, amidst uncertain terrain. But it made me sick post election that I ignored the sub-headline from a link I had posted often, "Post 9/11 Parity":http://people-press.org/reports/display.php3?PageID=750

In my '04 Excel model I used +2 party ID toward Democrats in the presidential race. And 62% toward Kerry among undecideds, which seemed conservative. That combo spit out Kerry as a a slight projected winner, by less than 1%. A few days after the election, when I adjusted to even party ID and Kerry with 4-6% less among late undecideds, the Excel model suddenly agreed with the national popular vote, Bush by roughly 2.5%. That's the main reason I'm convinced the '04 vote count was legit, though we were obviously robbed of tens of thousands of votes in Ohio and elsewhere due to low tech suppression.

Right now I think the party ID is artificially tilting our way by a point or point in a half due to GOP implosion. If Bush were closer to 50% approval rating I would use maybe +1.8 to +2.3 party ID edge for Democrats. That's my best guess where the nation stands right now. But very tough to judge since the Republicans still have post 9/11 advantages in federal races due to terrorism fear.

Yesterday I was posting in disbelief while looking at the Strategic Vision surveys, indicating 71-74% pf responders in every state expect a homeland terrorist strike within six months. I'm still reeling from that and may adjust the party ID downward in federal races. That impacts competitive senate races in particular. In governor races the party ID doesn;t need to be adjusted and we're in great shape. I just wish there were more emphasis on gov races here. In early November I could easily see another scenario where we're winning gov races nationwide, but DUers are bitching about narrowly lost and therefore stolen senate races.

That's quite a mouthful for me to chew. I'll have to read it a few times and reflect on it. One thing I'm wondering is, if we should just forget about weighting and select respondents randomly. I suppose you would have to survey a lot of people for that to work out.

Here is a chart of party identification, up to 2004:

http://www.umich.edu/~nes/nesguide/toptable/tab2a_2.htm

This is an example of why I am skeptical of the use of equal weighting in polls.

Check out this DU Journal article about Iraq war polling, which is currently on the front page:

http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=364x1540969

Conclusion: People are stupid. In addition, the persistent widespread public belief in Administration lies about the invasion of Iraq is a symptom of our complicit news media.

I have learned from your contributions to this message thread. Thanks.

Lasher

Lasher, that ANES table is another interesting link I hadn't seen before. With just a quick glance you can see the correlation between our worst years and the party ID margin. In '94 and '02 it was "only" 6 points, based on the criteria they use, while our best years are double digits. Should be promising for this year since they list '04 as a +8. It's obvious we have improved our party ID standing since '04 by at least 2-3 points.

I've never seen grade schoolers and high schoolers listed in an overall study like that. One thing jumped out at me from that table immediately but I can't guess the reason: percentage of grade schoolers identifying themselves as independent soared to 25% in '04, far above any previous number. Our figure dropped only 5% from that grade schooler category from '02 to '04 but Republicans plunged 11%. So the kiddies abandoned Bush and the Republicans and shifted to thinking independent. Does that have any meaning at all? Are the parents telling them something? Most likely I should just ignore it.

I didn't notice the shift until you pointed it out. I don't think those grade schoolers are identifying themselves as independent because they picked it up from video games, and they're getting the idea from somewhere. Guessing that most parents of grade schoolers are younger than 40, you might want to look for a related shift of voters in that age group.

After all these years of trying to get control of our government, the GOP finally got it and their collosal failure in every aspect has of course been met with disappointment. It is logical to suppose that you have detected a reaction to that. I sure wish we had a similar source of information that covered '05 and the first half of '06, when dissatisfaction with the GOP increased sharply.

There were even major full page ads from liberal groups with headlines like, "Galluping to the Right."

There was plenty of evidence Gallup turned toward the right and was improperly weighting its polls. Many of them had significantly more Republicans than Democrats in the sample but the numbers were released without highlighting that.

DUers wanted the typical +3 to +4 party ID edge toward Democrats in pre-election polls, since that had been the exit poll norm in previous presidential cycles. In '04 it turned out to be even party ID at 37-37, so the adjustments we wanted in many of the major polls would have been wrong. That's if you believe the 37-37, of course. I do, based on evidence before the election from PEW and others, but I realize many DUers insist the party ID was doctored along with the exit polls.

The pollsters who came closest to the final result in '04 had the party ID basically even and didn't allocate a heavy percentage of the undecideds to Kerry as the challenger. I think they guessed right, more than anything brilliant. Rasmussen was in that group. He's certainly a right wing pollster. So is Strategic Vision. Survey USA has been lumped there as well. Maybe only Zogby among the biggies is considered a left leaning pollster.

Most pollsters have personal leanings towards one party or another, and for the most part, their personal idealogy doesn't affect the results of their polling. The primary difference between pollsters is the methodology they use, which vary from agency to agency but are consistent within one agency. In some cases, a pollster like Zogby may be most accurate. In other cases, it might be Rasmussen, and in other cases it may be Strategic Vision and in other cases, it might be SurveyUSA. The most important thing to do, therefore, is to look at all the polls all the time, and instead of relying on one agency or one methodology, look at all the polls and pick out the trends.

For instance, in New Jersey, you can find recent polls that have Kean up by four points and you can find recent polls that have Menendez up by eight. No one can be sure which is more accurate, but looking at the way all the polls have trended, Menendez's numbers are on the rise as undecideds make up their mind. That's good news for us, regardless of what one or two polls may say. On the other hand, Strategic Vision is one of the only pollsters covering the Washington race right now. They have Cantwell's numbers decreasing steadily. Some people might think that since Strategic Vision is a Republican polling group, they may fluff their numbers. Even if this is true, their methods for publishing polls are consistent, and even if the numbers are off by five points, Cantwell is still in a slide, and that's something to worry about.

In my home state of Virginia, Zogby shows Webb down by five points while SurveyUSA shows Webb down by 17 and Rasmussen by 10. Which is most accurate? I don't know, but consider that before the primary, all polling had Webb down by 10 to 15 points, and Zogby had him down only by 7. This shows that Zogby consistently has shown this race to be more close than the majority of other polling agencies. Zogby may actually be the most accurate--no one knows for sure--but it can be assumed that a Zogby poll will show a more competitive race in Virginia than a Rasmussen poll. That helps us look at the big picture and look at the trends when individual polls come out.

Resident Apologist, I will happily elect you.

two steps forward, one step back.

TC

But I think I told it like it is, at least if you assume the latest polls to be 100% accurate. Don't dispair, I think we'll win in Ohio. Less than a week ago I was high-fiving with the Buckeye Democrats over the gains shown in the Zogby poll.

SurveyUSA has Allen burying Webb, 56-37.

I doubt it's that decisive, but the 5 point gap from another recent poll was probably much too kind to Webb. That was my guess in several threads. Likely a 10 point race if it went to the polls today.

Repukes have been attacking Zogby for their latest Virginia poll that showed Webb behind by only 10. Here's a link to Survey USA, where you can view the poll in question:

http://www.surveyusa.com/electionpolls.aspx

That's OK Virginia Democrats, keep on doing what you've been doing!

Rasmussen had Allen by 10, 51-41.

Zobgy had Allen by 5, 44-49.

Survey USA has Allen by 17, 56-37.


Zogby was also the poll that had Webb down by seven back in March, before he won the primary, which was an outlier to every other poll conducted on the race. I don't know what questions Zogby is asking, but his numbers consistently show more competitive races than the vast majority of other pollsters.

Kean was leading because of his name recognition from daddy and the perception that Menendez is part of the corrupt democratic machine. Menendez slaughtered Kean in the debate and the polls are reflecting that.

Sounds like they're ready for Whitewell to count the votes.

That said, I don't think Brown is a very good candidate.

Besides TheVirginian has assured us everything is legitimate. That is good enough for me. :)

Why is that a problem for you? Is there something in his Congressional voting record or stand on the issues you don't like? Can you be more specific?