Google Ads


Site Stats

KS-04: SUSA Says Pompeo Leads by 10, Goyle Poll Claims Otherwise

by: James L.

Thu Sep 16, 2010 at 7:57 PM EDT


SurveyUSA for KWCH-TV (9/14-15, likely voters, 8/9-11 in parens):

Raj Goyle (D): 40 (42)
Mike Pompeo (R): 50 (49)
David Moffett (L): 3 (4)
Susan Ducey (RP): 4 (1)
Undecided: 4 (5)
(MoE: ±4.3%)

SUSA strikes again, finding Dem Raj Goyle trailing Mike Pompeo by 10 points in this deep red district. Pompeo is helped by an implausible 66-22 lead among 18-to-34 year-olds (up from 48-26 in August). Eventually, SUSA will have to seriously examine why their methodology is conducive to this problem while other IVR pollsters, like PPP, haven't been plagued by the same issue. Another issue is a bit of procedural sloppiness - SUSA tested Libertarian David Moffett, despite the fact that he dropped out of the race and was substituted for Shawn Smith a week ago. Meanwhile, the Goyle campaign has released their latest internal poll:

Gerstein | Agne for Raj Goyle (9/8-9, likely voters, 8/10-11 in parens):

Raj Goyle (D): 44 (47)
Mike Pompeo (R): 46 (50)
Shawn Smith (L): 4 (-)
Susan Ducey (RP): 2 (-)
Undecided: 4 (3)
(MoE: ±4.4%)

Their full polling memo is available below the fold. One area of concurrence between SUSA and Goyle's pollster is that Goyle is running quite strongly among independents: SUSA gave Goyle a 7-point lead among that bloc, while Gerstein has Goyle up by 8.

James L. :: KS-04: SUSA Says Pompeo Leads by 10, Goyle Poll Claims Otherwise
Tags: , , , , , , , (All Tags)
Print Friendly View Send As Email

I'm afraid I'm still not seeing this
Particularly with Brownback and Moran coattails.

Unfortunately
Unfortunately, I have to second this. Is Pompeo that detestable?

[ Parent ]
There may or may not be coattails in a partisan blowout. Remember New York......
In New York there's so much discussion here that with Democratic blowout wins at the top of the ticket in a strongly Democratic state where Democrats always win everything anyway, Democrats might not be so motivated to show up.  That's plausible, but it's hard to gauge.

But the same exact thing is equally plausible in Kansas.   Yes Brownback/Moran could have coattails, but the flip side is that Republicans always win everything in Kansas, so what is there to motivate them?  The only thing I can think of is that Democrats have in fact controlled the Governorship for 8 years, but I don't get the sense that getting the governorship back is making KS GOP voters into eager beavers more than in any other election.  As for the Senate race, no Democrat has held a Kansas U.S. Senate seat since the Great Depression, so no motivation there.

Plus, I could see KS-04 Democrats, if properly informed, learning of the opportunity here to make a splash and having some outsized motivation to turn out in spite of the anti-Democratic wave nationally.  Again, the same as I can see the utter demoralization of New York Republicans actually motivating pockets of voters in NY-29 or other U.S. House districts where there's real opportunity to take back something.

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


[ Parent ]
Goyle's lucky it's Moran and not Tiahrt
I'm also ambivalent as to whether there will be coattails or not in the non-competitive races in NY and KS, but if there are, they would be stronger in KS-04 with native son Tiahrt instead of Moran atop the ticket.

20, CD MA-03/NH-01/MA-08

[ Parent ]
Not sure if it applies here...
We had this discussion about New York two days ago and I would agree its plausible in Kansas too.  I have more reservations about it in Kansas though because it is Republicans dominating a Republican state in a Republican year where Republicans want to go out and give a piece of their mind at the polls.  

We might actually see the opposite in a place like Kansas where the Democrats are generally demoralized and see the Republicans running away with it anyway.  We might see something similar in Pennsylvania and Michigan where the Democrats are not fired up and their clocks are already getting cleaned.  Come November, they might see Corbett in PA and Snyder in MI up by double digits and say there is nothing to vote for.  I think that happened to the Republicans in both states in 2006 and 2008.

23, male, center-right cynical Republican, PA-7


[ Parent ]
I don't think it's valid to compare Kansas with PA or MI. Kansas is different because...
...Democrats are used to losing and likely are keener about paying attention to opportunities to win wherever and whenever they arise.

Democrats in MI and PA are used to winning, so they're particularly demoralized in a bad year.

But when losing is your norm, the psychology is different.  Turnout is not going to be any worse in Kansas this year than in most years.

For Kansas Republicans, you may or may not be right, but your hypothesis is arguable at best.  The same psychological motivation just isn't there when you're used to winning all the time.  In 1994 the wave hit Kansas because there were a couple Democratic Congressmen there to knock out.  But Moore's seat is all there is, he's retiring and it's already a near-certain pickup, and that's not going to affect KS-04 anyway.  If nothing else I guarantee any elevation in GOP turnout in Kansas compared to the state's norm is a LOT LESS than the elevation over the norm in GOP turnout in swing states or blue states.

Goyle has a real opportunity if he can get district Democrats to wake up and see what's happening.  He's smart to release private polling, as it might help toward that end.  It's a longshot, yes, but the opportunity for an upset is real.

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


[ Parent ]
I don't think Brownback has coattails
I sense very little enthusiasm for Brownback among Kansas Republicans. I get the sense that they are uncomfortable with him coming from Washington back to run Kansas. This is an anti-Washington year and he is coming from Washington. That doesn't mean Republicans won't vote for him, just that they aren't enthusiastic about it. I think Democrats in Kansas are more enthusiastic to vote against Brownback than Republicans are to vote for Brownback. That's the truth. You hear a lot more anti-Brownback sentiment than you do pro-Brownback sentiment.

[ Parent ]
internal
Goyle's internal has him down by 2. If you knock off the standard 5 for selective release of internals, he's down by 7, which makes this a rare case where the Survey USA poll basically "agrees" with other polls.

41, Ind, CA-05

The "standard 5"
Not sure how 5 became "standard".

Two years ago, Glenn Nye was selectively releasing internals showing him down by five points (twice) -- in mid-October. Was he really down by 10?

I'm really reluctant to apply a rule of thumb like that in such a blanket-like manner.


[ Parent ]
Internal polls
Aren't internal polls typically very expensive because they're accurate?

[ Parent ]
The internal polls that campaigns use for their own purposes are very expensive (and accurate), yes.
However, there's no telling whether any particular internal poll is one of these expensive, accurate ones, or one that the campaign had the pollster cook up to release to the media.

[ Parent ]
I disagree
As Dana was saying the other day, a lot of the better-known pollsters (D & R) have reputations to protect - shooting out junk can hurt them with future clients. That's especially true for those pollsters who have corporate clients (probably where most of their money is made), and for whom accuracy is all that matters (since they polling never sees the light of day).

[ Parent ]
But David, that doesn't mean campaigns don't cherry-pick what polling data they release, which they do......
Campaigns can and do release "informed ballot" numbers at times, and they won't necessarily disclose that's what they're releasing.  As long as they can claim the data is legit on its own terms, the pollster's reputation can remain intact, even if the campaign's release is misleading by disclosing only "informed ballot" numbers or has a "target" turnout model (i.e., the model they think they need to win as opposed to a less favorable model that nonpartisan observers think is more likely to appear on election day).

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

[ Parent ]
Sure, campaigns can try all kinds of shit
But a reputable pollster would be furious if a campaign tried to do this. It would be a good way to ensure that your pollster never did any work for you again.

Put another way: I've never heard of this happening, not with a real pollster, anyway. And while I realize that good pollsters would probably not want to dime out even shady clients, there would be rumors. And I've never even heard a rumor of something like this.


[ Parent ]
My two cents here
I agree with David and pretty much everyone else that these pollsters have to be accurate to stay in business, and that the more prolific ones (GQR, Tarrance, etc.) are probably very good at what they do. Yet most internal polls don't get released, and the skeptic in me leads me to believe that the ones that do get released are sometimes: a. push-polled or b. outliers that surprise even the candidate. That's why we saw, say, Steve Southerland in FL-02 tout an internal with him up 52-37, but have never seen another internal from his camp.

I respect internal polls and love seeing them, but I take them with a grain of salt, particularly from challengers. And sometimes, they can be dead wrong--Clay Shaw said after his 5-point loss to Ron Klein in '06 that he was up 5 in his internals right up to the end.

20, CD MA-03/NH-01/MA-08


[ Parent ]
About released polls....
I've never had the impression that the unrealistic internal polls were "cooked". Rather, they are just the outliers that occur every now and then that are lucky for the campaign. In other words, after weeks of sorting through average or bad internal polls, one of their polls shows the candidate 10 points ahead... wow, let's make it public!

That said, that's my guess coming from someone who has never had a job in a political campaign, so maybe there is some book cookage going on.


[ Parent ]
this
This is the impression I have, that they take a lot of polls and release the ones where they got a lucky sample. That said, I have never worked on a campaign and have no inside knowledge.

41, Ind, CA-05

[ Parent ]
what is the deal with Survey USA?
That is weird to have so many polls showing Republicans up big with young voters.

What kinds of young people
1.  Stay home on Friday nights to wait for a robocall
2.  Are the ones that even answer a poll
3.  Use landlines
???
SUSA numbers

I mean, really, what kind of young people are answering these phone calls.  18-34 is a big group, but a lot of people don't bother with land lines anymore, and the 18-21 bracket is almost certainly never getting polled accurately since they're either in college and use a cell phone or not the head of the household.  What's left over may well be a lot more conservative than the demographic in general.


[ Parent ]
the weird thing is
SUSA did some polls a while back I think where they included cellphones and it didn't change their bottomlines, young voters were still overwhelmingly Republican according to them.  

[ Parent ]
Hard to believe.
Link?

[ Parent ]
Exactly.
"More than a third of people under age 35 - including about half of those age 25 to 29 - have only cell phones."   http://www.msnbc.msn.com/id/37...

[ Parent ]
There needs to be a big banner put up over every political website
saying "Don't trust crosstabs." It's ridiculous to try and draw conclusions from subsamples with huge margins of error.

[ Parent ]
Yup, that's correct. Virtually the only crosstabs I give any trust are...
...for white voters, since in almost any poll, and definitely in all statewide ones outside Hawaii, white voters make up at least two-thirds of the sample and in most states over 80%.  So the margin of error on that subsample is a lot lower than all other demographics.  Crosstabs on gender also can be OK if the total sample is large enough, since by sex each group is roughly half.

Age crosstabs are far and away the most useless, because in most polls they break down age by 3 or 4 categories, and youth vote crosstabs are the worst since they make up only a few dozen survey responses.

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


[ Parent ]
still
shouldnt this margin of error be even in both directions. it's always GOP favorable in this demographic.  by this logic, sometimes it should be 90-10 Dem, right?

18, Dem, CA-14 (home) CA-09 (college, next year). social libertarian, economic liberal, fiscal conservative.   Everybody should put age and CD here. :)

[ Parent ]
Well it should be random, not necessarily "even," but are you so sure about...
...SUSA's young respondents consistently breaking heavily Republican this year?  So many people here say that, but has anyone here really studied the data to prove it?  The alternative is people are just hunting for funky crosstabs in unfavorable polls and noticing this one whenever it occurs, and developing a perception of a pattern that might not exist.

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

[ Parent ]
You are totally welcome
To comb through the archives and crunch the data.

However, I do feel the need to say that I resent your suggestion that we trot out the 18 to 34 year-old issue as a convenient way to dismiss unfavorable polls. The fact is, it is a pretty glaring issue for SUSA this cycle, and we've even highlighted that trend in cases where the toplines were good for Democrats.

And when we don't talk about it, it's often because I'd rather not sound like a broken fucking record!


[ Parent ]
Looked through
There seems to be a pattern.  Not universally, but more often than not that their polls have high Republican favorables.

On a side note, I only have a cell phone and it has a western Pennsylvania number.  I got a robocall poll today and it was not even for the right district as my cell phone exchange number is from.  The district they were polling for was roughly 30 miles away (PA-4) from where landlines with similar numbers are located (PA-12 or PA-18).  Cell phone polling is bound to be highly inaccurate because people are moving cell phones across area codes now.  

23, male, center-right cynical Republican, PA-7


[ Parent ]
Hey, I'm a Democrat, I'm on your side! And I didn't know SSP moderators shared....
...the suspicions about SUSA's age crosstabs.  I see it in comments from various SSPers all the time, and I see your frontpage posts occasionally mention it, but I don't take mere mention as clearly sharing suspicions about SUSA's age crosstabs.

Further, I think we're all subject to the temptation to dismiss unfavorable polls in one way or another.  That's not necessarily unhealthy, it's part of democracy and politics for people to pick apart the opposition in one way or another as a way of forcing a reality check.  But it's also emotional self-medication we all engage in to try to belittle whatever seems to be bad for our own side politically.  It's human nature.  I do it, too.  I try to catch myself, but I've picked apart unfavorable polls a bunch, sometimes justifiably, but sometimes only to find my critiques are unwarranted.

Regarding age crosstabs specifically, I stand by it that it's hard to put too much stock in them because young voters most of the time are among the smaller subsamples.

And to be clear, I'm not defending SUSA at all because t's clear they're producing a lot of ridiculous outliers in their toplines.  I'm just not sure that their young voter sampling is really the problem, rather than methodological problems across their entire samples.  After all, for most of their polls, changing the numbers for young voters to make them favorable to Democrats doesn't actually change the toplines much, since young voters are a small percentage of the samples.  Even if one were to increase the young adult vote share to align with the 2006 midterm exit polling along with saying most of them vote Democratic, still the toplines don't really change that much in these SUSA polls that show overpowering GOP strength in the toplines.

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


[ Parent ]
Oh, and James, let me add....
...to the extent you took my post as critical of you or others, I apologize.  I'm not challenging anyone's integrity on this point.  But I do know that in self-reflection I've at times been guilty of errantly seeing things through blue glasses.  Ever since Kerry lost to Dubya, which shocked me at the time, I've tried to be more introspective and also examine my fellow Democrats' assertions and arguments with more scrutiny, for the sake of staying reality-based.

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

[ Parent ]
That's fair
I also apologize if my response was overly reactive - it came at the end of a long day. I will submit, though, that SSP tries hard not to massage the truth to fit a worldview wears Dem-tinted glasses. In other words, I think we're sufficiently pessimistic!

As for SUSA... in November, we'll learn a lot about the accuracy of these polls.


[ Parent ]
Oh, trust me, you and the other moderators never overstate our chances! In fact...
...you guys were rating all the races over the spring and summer more pessimistically than I did.

I'm come around to surrendering the House mentally, and overall I no longer question your team's ratings at all.

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


[ Parent ]
Um no
There needs to be a default where people can't type the nonsense sentence of "don't trust crosstabs".

Could we finally put that silly sentence to rest?  Crosstabs on a single poll can always be questioned.  Crosstabs that show consistent behavoir over and over again represent not just a small subsample.  That's like saying you can't draw conclusions from one individual voters behavoir... even if you have data from millions of these individuals.

It's utterly ridiculous to ignore SUSA's consistent issues with young voters.  Either they are worong or they are right about this phenomenon.  "ignore it" is no longer an option for anyone with cumulative memory ability.


[ Parent ]
Gerstein/Agne
hit the nail on the head in their analysis. If you want to know what's going on in KS-04, why Goyle is running competitively, read their analysis. They nailed it.

I just started seeing Pompeo's general election ads a few days ago. Generally speaking, Goyle has had more TV ads than Pompeo up until now. That is part of what has helped establish Goyle's name recognition and high favorables. Goyle has a lot of money, but so does Pompeo.

All of Goyle's ads have been positive. He talks about saving American jobs, supporting veterans, and opposing the Wall Street bailout. Goyle is running toward the middle, but Pompeo supporters are trying to portray Goyle as a far-left liberal.

Goyle has not had many negative ads run against him yet, maybe none. I have not seen any. Pompeo is just starting a major ad campaign, which I am sure will eventually include a lot of negative ads against Goyle. The longer the race remains competitive, the more threatened Pompeo will feel and the more negative ads he will run.

It would not surprise me if negative ads from Pompeo start hurting Goyle's high favorables and reducing his support among moderates. The impact of any negative ads probably will not show up in polls until October. Pompeo has enough money to run negative ads around the clock and it's just a matter of time before he does it. I hope Goyle stays positive, but he will have to respond in some way.

Pompeo and Goyle are scheduled to debate next week.

As for SurveyUSA, I have always thought there was something amiss about their crosstabs.


DONATE!
www.goyleforcongress.com

22, male, VA-10


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

SSP Resources

Blogroll

Powered by: SoapBlox