Google Ads


Site Stats

WI-Sen: Suprisingly Close Race for Feingold

by: Crisitunity

Tue Jun 29, 2010 at 1:40 PM EDT


PPP (pdf) (6/26-27, Wisconsin voters, 3/20-21 in parens):

Russ Feingold (D-inc): 45
Ron Johnson (R): 43
Undecided: 12

Russ Feingold (D-inc): 45 (48)
Dave Westlake (R): 38 (31)
Undecided: 17 (21)
(MoE: ±3.9%)

There was a general sense of Russ Feingold having dodged a bullet when ex-Gov. Tommy Thompson decided not to run (in PPP's March poll, Feingold led Tommy Thompson 47-44). In a bit of a surprise, though, likely GOP nominee Ron Johnson performs about as well as Thompson, trailing Feingold only 45-43. PPP's Tom Jensen speculates Republicans may have actually done themselves a favor here by running a fresh face (Johnson) instead of the stale Thompson; with only 20/18 favorables right now, Johnson does certainly have a lot of upside. (Feingold's approval is 42/42.) Of course, on the other hand, some of Johnson's support now may simply be because he's something new and different, and while Thompson had some moderate crossover appeal, the very conservative Johnson may not have much of that once the candidates start talking about specifics.

PPP may have run into a more conservative batch this time than last time; today's sample broke 48 Obama, 47 McCain, and it's also apparent in the trendlines for the low-profile Some Dude in the race, Dave Westlake. Regardless, it's a pretty clear signal that Russ Feingold (and the DSCC, unfortunately) are going to have to fight this one out.
Crisitunity :: WI-Sen: Suprisingly Close Race for Feingold
Tags: , , , , , , , (All Tags)
Print Friendly View Send As Email

WI-Sen: Suprisingly Close Race for Feingold | 39 comments
Looks like Research 2000
is going the way of Strategic Vision http://www.dailykos.com/story/...

And I'm not sure that the sample is really only D+1 in Wisconsin...

28, Liberal Democrat, CA-26

I didn't know where else to comment on this
but it's some pretty big, shocking news. Possible lawsuit here, as it looks like the daily tracking numbers were pure and simply fabricated when analyzed statistically.
[ Parent ]
I'll read it later
Hopefully Nate chimes in too. Very disturbing though to say the least.
[ Parent ]
On cue from Nate
http://www.fivethirtyeight.com...
Following the results of an investigation  conducted by three independent researchers, Mark Grebner, Michael Weissman, and Jonathan Weissman (Michael Weismann assisted FiveThirtyEight in its investigation of another allegedly fraudulent pollster, Strategic Vision), Daily Kos will be suing its former pollster, Research 2000, which it dismissed  three weeks ago, partly as a result of its poor results in our pollster ratings as well as other concerns  I had expressed to Markos Moulitsas about their polling.

[ Parent ]
should have bolded the
Daily Kos will be suing its former pollster, Research 2000

[ Parent ]
Ho. Ly. Shit. I never would have guessed R2K...
...is a fraudster.

Strategic Vision always "conducted" its "polls" for no apparent client, as far as I could tell, and published them on their web site presumably for the publicity.  So no real shock on them.

But R2K is a major trusted media pollster.  Del Ali is quoted everywhere, all the time.

Why would an outfit like that run fraudulent numbers?  Was it to save money and pass on the "savings" of not polling at all to the clients to keep business alive?  That's all I can speculate.

This is a big deal in polling.

43, male, Indian-American, Democrat, VA-10

[ Parent ]
I
am not a lawyer but it sounds like the kos has a really good case here, am I correct in this thinking? I knew they were bad but actually fabricating numbers, that is sick, really sick.  

Proud member of the Indiana Democratic Party from IN-9.  
[ Parent ]
Depends if R2K has proof they did the polling
Else it means finding their own stats experts to rebut and/or bamboozle.
[ Parent ]
If they did actual polling, proof should be very easy since it's live caller......
If they did the polling, then there are real live human beings out there who can testify they called people, asked them a list of questions, took down the answers, and turned in the data after finishing the calls.

If R2K is faking polls, the above evidence probably doesn't exist, because the main value of faking is simply to save money.  Much of the cost in live caller polling is labor, so that's what's saved in faking data.

43, male, Indian-American, Democrat, VA-10

[ Parent ]
Feingold
He never wins by big margins.  In '98, he beat Mark Neuman by 3% after trailing him for much of the fall.  I remain confident that he'll win this one too.
R2K responds
R2K president Del Ali tells TPMmuckraker in an email, "I have much to say, however, I am following my attorney Richard Beckler' ESQ's counsel and referring all questions to him. I will tell you unequivocally that we conducted EVERY poll properly for the Daily Kos."  
I dunno,
the statistics look pretty damn definitive here, unless R2K was making some really funky adjustments which they haven't released yet.  
[ Parent ]
Dunno
I'll leave it to the pro-stats guys to fight it out but I will say it reminds me quite a bit of Rasmussen daily presidential tracking from 2008 which at times looked suspiciously consistent day-to-day, especially compared to the way Gallup fluctuated so much.
[ Parent ]
Funny how the vast majority of the comments are about
the DKos pollster story.
In terms of Wisconsin, I am surprised it is this close. Could it be possible that Ras was right?  
Rasmussen is often right
He is often very wrong. On average about five points GOP leaning. An average is exactly that.
[ Parent ]
Feingold will have more trouble
than he really should because he has never been very popular in Wisconsin, not like the much quieter, and yet just as liberal, Herb Kohl.  
[ Parent ]
Another ridiculous PPP sample
Even worse than the Pennsylvania sample.

Obama carried Wisconsin by 14%, but in this sample it is Obama by 1%.

Um, heah, the race is close if you think Mccain turnout will stay the same and Obama turnout will go down 28%. That is about the equivalent of every Obama voter under 40 not showing up, plus every African American voter over 40.

Remember these numbers don't reflect changes in votes from Obama to Mccain or D or R, or vice versa.  They are merely guessing that turnout for Mccain voters will X% of 2008 while Obama turnout will be (X - 30%).

AA turnout in Wisconsin was only 5% total in 2008, and Wisconsin has traditionally high turnout, so these number for this state are beyond a joke.  PPP should be ashamed to put out such garbage.

Bottom line, Feingold is very close to safe Dem if a sample like this shows him leading.

I really don't see how this is so hard to believe
Considering Wisconsin was a 50/50 state in both 2000 and 2004. Obama's margin in 2008 was exaggerated by McCain giving up there.
[ Parent ]
But that isn't what they are saying!
Look, I can believe 50/50 just fine, but that isn't what they are saying.

They are saying turnout for Mccain voters will be X%... whatever the droppoff between a presidential year and off year is.

But then the Obama turnout will be x - 30% more.

When Wisconsin was 50/50 this was because a lot of those Obama voters voted for the Republican, not because they were vaporized into thin air.

[ Parent ]
What's the difference?
50/50 v. Obama winning in '08 and D turnout going down?

It sounds like you're making a mountain out of a rumble strip based on a personal animus towards PPP.

If not, where are your facts and data suggesting that PPP is wrong? Show us how many Obama voters will turn out -- and still favor Ds. What credible poll backs up your assertions?

[ Parent ]
Oh please, the difference is obvious
I have no animus toward PPP so stop pretending I do.

PPP has 25-30% of the Obama voters vaporized... not changing their votes to Republican, just not existing.  That's absurd.

For every 100% unit of Mccain voters, only 72% of Obama voters are included in that poll.  This is like wiping out more than every Obama voter from Milwaukee, Racine, Kenosha and Ozaukee counties.

This can be put in perspective in several ways.  One is to say only 5% of Wisconsin votes were African American.  If zero African Americans were to vote, that would only decrease the number of Obama voters by 10%, nowhere close to 28%.

Then percentage difference between 2008 turnout and 2006 is 2006 is about 28% less.  So PPP is projecting Obama voters turn out at a normal off year percentage, while Mccain voters turn out at 100% of the presidential year.

Finally, AA turnout was 5% in 2006 and 5% in 2008, so presuming it would be lower than even 4% is illogical.

I don't know what turnout will be, but for Wisconsin, I think it is reasonable to presume that for every 100% Mccain unit that turns out that the corresponding Obama unit will turn out at 95%.  I suspect it will be less than that in Wisconsin given both the relatively tiny AA percentage of the voters, and their historic turnout patterns.

In contrast, Virginia AA turnout in 2008 was 20%, and that dropped to 15% in 2009 (and AA's voted more Republican in 2009).  Better than 10% of Obama's Virginia voters did vaporize...

So I wouldn't say Obama voters will turnout 95% of Mccain voters in every state, but Wisconsin is only 5% AA, and Pennyslvania between 5 and 13%.  In some heavily first-time AA voting areas (Virginia, Carolinas, Georgia) Obama voters likely will turnout with a similar dropoff as Virginia ones in 2009, but there is no way to suppose Mccain voters will turn out at 100% rates while non-AA Obama ones will turnout far, far worse percentages than they did in Virginia.

To oversimplify it, I could take WI or Penn polls that vaporize 5 to 10% of Obama voters, but not 22 and 28%.  It's absurd.

(Again, this is completely different than Obama or Mccain voters changing their minds.  This poll simply presumes that numbers of Obama voters equal to those in Milwaukee and those other counties will be not voting at all.)

[ Parent ]
In MA,
which has a decent-sized minority population but not as big as PA, Brown received about as many votes as McCain while Coakley received far fewer than Obama. But the one point of yours that I agree with is that Wisconsin is a very high-turnout state.

21, dude, RI-01 (registered) IL-01 (college)
please help Japan. click "donate funds" in upper right and then "Japan Earthquake and Pacific Tsunami." http://www.redcross.org/
[ Parent ]
Coakley was an abysmal candidate
though I know most people didn't agree at the beginning.

The PPP poll is an epic fail in several ways, but looking at turnout is a huge one.  They are projecting the worst turnout in modern history for Wisconsin, by far.

At the same time, they are projecting a turn toward the Democrats in the turnout that does occur... from Obama +1 to Feingold +2.5!  

28% of the Obama voters choose not to vote, but Feingold improves on Obama's performance among those who do vote.  This would mean that an actual random sample of 2008 voters would have Feingold between +15 and +20.

It is pure guesswork to suggest that in Wisconsin we'll lose even 5% of the Obamas in comparison to the Mccains, but it is completely absurd to postulate a 28% Obama voter loss, if only because that would require turnout to be low to an impossible degree.

[ Parent ]
R voters are more enthusiastic
D voters are not.

I think this comment from user DCCyclone (title bolded) is the best basic refutation of what you say http://www.swingstateproject.c...

  tommypaine, do you understand the concept that the same people don't vote?......
Election to election, different people vote.  Some of that is normal turnover from old people dying and young people voting for the first time.  But much more of it is that tens of millions of voters are interested in voting only when the Presidency is on the ballot, and won't show up otherwise.  And the political reality in America is that those people are disproportionately Democratic voters.

So in a midterm, the electorate is normally more Republican than in a Presidential election.

And in this midterm, Republicans are expected to vote in higher numbers than is normal even for a midterm.  That is because they hate Obama and Democrats have supermajorities in both chambers of Congress, and our party favors policies that Republicans hate.


[ Parent ]
Duh, hello, obviously
Are you seriously not getting it?

There are variations in turnout every election, but the PPP sample is not a variation.  It postulates that turnout will be historically low... that for 100% units of Republicans only 72% of the parallel unit of Democrats turnout... that of the turnout that does turnout it will vote more democratic than 2008...

Why can't you understand this is a question of degrees?  Please list a mideterm eelction where anything like this has happened.  Obviously you can't.

So just for laughs, what is your number?  If 100% of Mccain voters turned out, what would be the Obama percentage?  I'm willing to accept in Wisconsin a 5% drop relative to Republicans... so in broad numbers in Novmember 75% of the Mccain voters show up, and 70% of the Obama voters.  That would refelct about a 2.5% lower turnout than normal.

In real numbers, it means Mccain turnout is normal for a midterm, while turnout for Obamas is 65,000 less than normal... about the equivalent of the entire Brown County (Green Bay) Democratic vote in 2008.

Try to think of that, I'm willing to accept the equivalent of zero Obama voters from the fourth largest country in the state.

5% is a massive shift, but I'll give you that.  PPP instead wants to remove 28% of the Obamas on top of the normal difference between a presidential year and a midterm.  So what is your number?

Either you just have some kneejerk reason to cheerlead for PPP, or you haven't thought this through very well.  And before you did yourself in deeper, take a look at PPP's Ohio sample, which is about 1% different than the 2008 numbers.

In other words PPP is vaporizing less than 2% of the Obamas in Ohio, but 28% in Wisconsin.  Care to explain how that is something other than looney?

[ Parent ]
Despite your personal insults
I'll try to answer your questions seriously.

First, the numbers in your 2nd paragraph are faulty, as approvals change. An election where such happened is '94.

Second, a 50-50 electorate in WI would simply be a move back to its norm as a purple state. '08 was a great election for us in many states, including Wi. Unfortunately, a move back to a 50-50 equilibrium is not loony.

Third, the numbers associated with your 4th paragraph are based on no changes w/r/t independents.

Fourth, in the future, I'd appreciate the same level of  respect that I'm trying to give you. Personal insults do not promote rational discussion.

[ Parent ]
Please
You deliberately mistated my opinion several times.  Just stop it and let's move on.

First, no it didn't.  PPP is postulating a Wisconsin turnout about 15% below the 2006 off year election.  Do you agree?  Why?

Second, perhaps you are simply not understanding.  This has nothing to do with Wisconsin moving to a 50/50 state.  This is only relating to how 28% of the 100% of the Obama voters will not be able to vote for either a Dem or a Rep because PPP vaporizes them, while assuming a relative 100% of the Mccain voters can vote.  For about the tenth time, "changing" is not at issue here.

Now you need to take a step back and instead of focusing on what is not at issue here and what I have not said, and instead look at the poll and what I have written.

And then also read PPP's Ohio poll.  In Ohio they postulate a 2% turnout drop in Obama voters from 2008 to 2010, while postulaing a 28% change in Wisconsin.  These are not "changed" voters.  These are the numbers of who the voters did in fact vote for in 2008.

There is no way to reconcile 28% of Obama voters in wisconsin not voting this year, while only 2% of Obama voters don't vote this year in Ohio.  

That is what you are defending, that it will be 2% in Ohio and 28% in Wisconsin.  On Planet Outlier maybe, but here on Earth, those two things can not exist at the same time... and there is no way to logically justify such a massive difference between the two states.

[ Parent ]
Unfortunately
I suspect it is the Ohio poll that is the faulty one. Whatever the internals say I would tend to trust the topline horserace numbers. For example the final PPP poll of VA had McDonnell winning 56-42 with a McCain +1 sample. The exit poll had actual voters breaking for McCain by a margin of 51-43.

[ Parent ]
The personal insults continue
I state what I believe I see.

To suggest that

You deliberately mistated my opinion several times.

is libelous.

(Of course, if you have documentation of what's going on in my head, then truth is a defense against libel. In that case, I'm losing my mind. ;) )

Second, I acknowledge that I don't understand everything. Nevertheless, I see no difference between the two things you compare:

Wisconsin moving to a 50/50 state.

and your perception that
This is only relating to how 28% of the 100% of the Obama voters will not be able to vote for either a Dem or a Rep because PPP vaporizes them, while assuming a relative 100% of the Mccain voters can vote

In either case, the numbers still end up as 50-50. That's the topline result I expect for WI in a heavy R year such as many of us expect in 2010.

If this were '94, I believe that WI would lean R. But with the generic effect of demographics (I have not analyzed that detail), I believe that WI at 50-50 is a good sign.

If there are other polls that show a different situation, then I'll reassess.

I also believe that differences of opinion can be discussed without personal insults. I hope in the future that you can find a way to do so.

[ Parent ]
This was totally humorous to read through
hahahahahahaa, I cant believe neither of you gave up earlier
[ Parent ]
Jeezus will you cut it out already
"In either case, the numbers still end up as 50-50."

Clearly you aren't understanding the basic math here.  I've explained it to you several times.

I've written other numbers in the Ohio thread.  If you want to try and understand how one or the other of the PPP polls must be totally wrong, then take a look.

[ Parent ]
Duh hello
Are you really not getting it? EVERYONE here has tried to explain turnout to you. Everyone here knows you are wrong. YOU are the one not getting it, NOT us!  
[ Parent ]
Wisconsin, Pennsylvania, Iowa, Illinois...

[ Parent ]
I'm more interested in the party split
D 33, R 32, I 34 here compared to 39-33-29 in 2008. If lots of Republicans are calling themselves independent as I have postulated many times that would make sense. Especially since the baseline I've been using to calculate likely turnout is an average 6 point movement to the GOP, as it was in New Jersey.
[ Parent ]
You did the numbers and found it was 6%?
(Just checking because that's a good baseline to go off of.)
[ Parent ]
Going by the exit polls
Obama/McCain was D 44, R 28, I 28
Corzine/Christe was D 41, R 31, I 28

In other words a D+16 partisan electorate became D+10 in New Jersey between 2008 and 2009.

[ Parent ]
Currently thinking...
Dem - 35%
GOP - 35%
Indie - 30%

Feingold - 93/10/54 = 55%
Johnson - 7/90/46 = 45%

Still looks like a Lean Dem race to me, probably less competitive than WA-Sen. Keep in mind, in his '04 bid, Feingold siphoned 14% of Republicans and won 2-to-1 among Independents. I don't expect either of those in this environment, but I think he has some wiggle room.

For daily political commentary, visit me at http://polibeast.blogspot.com/ and http://twitter.com/polibeast

I got different numbers based on your model
Based on your model, Feingold would get 52%, Johnson would get 48%.  That's less wiggle room than I'm comfortable with.

40, male, Democrat, NC-04
[ Parent ]
The PPP sample has Feingold outperforming Obama
Solid 10 point win here.
[ Parent ]
WI-Sen: Suprisingly Close Race for Feingold | 39 comments

Copyright 2003-2010 Swing State Project LLC

Primary Sponsor

You're not running for second place. Is your website? See why Campaign Engine is ranked #1 in software and support among Progressive-only Internet firms. http://www.mediamezcla.com/

Menu

Make a New Account

Username:

Password:



Forget your username or password?

About the Site

Publisher: DavidNYC

Senior Editor: James L.

Senior Editor: Crisitunity

Contributing Editor: Jeffmd

Search the SSP archives:

Sign up to receive a daily e-mail whenever the SSP is updated:

SSP Resources

Presidential Vote by CD

SSP Document Collection

Race Tracker Wiki

Dave's Redistricting App

Blogroll

Daily Kos Elections
DemConWatch
Five Thirty Eight
Governing
Hotline On Call
Political Wire
Pollster.com
Prop*50
Public Policy Polling
The Thicket
Powered by: SoapBlox