Our Sponsors:

Read more »

Our Members

Many thanks to Stephen Coyne and Tara Mattina some of our many supporters.

ALL MEMBERS »

Why are the polls so wildly different in the Rossi-Murray race?

The answer lies in the assumptions pollsters make about the current party identifications in our state.

Dino Rossi

Dino Rossi Courtesy of Rossi campaign

I have worked in politics in Washington state in one capacity or another since 1982.  I have been a campaign staffer, a candidate, a party chairman, an office-holder, and a consultant, and I have never seen polls as inconsistent as the polls we are seeing in the current Murray-Rossi Senate race.  Some polls show Rossi slightly ahead, others show it dead even, and the Elway poll shows Senator Murray cruising to reelection. 

The Washington polling confusion is unusual enough that it is drawing national attention.  In releasing their latest results SurveyUSA included the following explanation:  "There is unique and in some ways unprecedented disagreement among pollsters as to where this contest stands, and anyone trying to digest these SurveyUSA results needs to know that within the past 72 hours, Elway has the Democrat up 15 points and Fox News has the Republican up 1 point. While it is possible to see 16 points of pollster disagreement on ballot measures and in low-turnout primaries, it is unusual to see so much variation in a high-profile, statewide general election…. The New York Times describes Washington state as the Bermuda Triangle of polling."

What's going on? According to conventional wisdom the difference lies in how the polls are being conducted, and that Murray has an advantage in live interviewer polls versus automated polls.  This is essentially the argument Stuart Elway offered for his outlier results.  Mark Blumenthal wrote about this thesis last week: "The live-interviewer pollsters, CNN/Time and Elway, show Murray doing consistently better than the automated polls from Rasmussen Reports, Fox News and Rasmussen subsidiary Pulse Opinion Research and SurveyUSA. This week, for example, two new Rasmussen and Fox/Pulse polls showed a very different result, Rossi leading Murray by 3 and 1-point margins respectively."

There are two propblems with this argument.  The first is not all polls using live interviewers are finding results consistent with the Elway and CNN polls.  The recent Fabrizio poll used live interviewers, and it showed Rossi ahead, 48-42.  In addition, I am aware of internal campaign polling done recently with live interviewers which yielded far different results than did the Elway and CNN poll.

Also, this discrepancy in live vs. automated polls is not showing up in other parts of the country.  In the Wisconsin Senate race, for instance, the latest CNN poll is virtually identical to the latest Rasmussen poll. 

A better explanation for all the variation, I believe, lies in how different pollsters screen for likely voters — not the method used to poll.  Specifically it lies in the assumptions pollsters are making about the partisan makeup of the 2010 electorate.

To be accurate, a poll must be done on a statistically valid sample of the population.  Your sample must have the right number of men and women, young and old, and the right geographic mix.  As a poll is being conducted, pollsters will work with their callers to get this right mix, or they will weight the results after the poll is done.  If you didn’t get enough responses from men, for instance, you can look at the responses from the men sampled, extrapolate, and adjust (or weight) the results accordingly.

This weighting is easy to do when you have known demographics, such as age. But what about trying to weight a sample to reflect party identification?  Party identification shifts over time, and no one really knows how many Republicans and Democrats are going to vote in each election (the so-called enthusiasm factor).  Quite often assumptions have to be made, and theses assumptions can dramatically affect results.

The question comes down to this: How large is the gap between Republicans and Democrats in Washington state in 2010?  The Elway sample included 39 percent Democrats, 26 percent Republicans — a 13 percent gap for the Democrats. At a time when the national generic ballot shows Republicans with an unprecedented advantage, is there still a double-digit party ID advantage here for the Democrats?


Like what you just read? Support high quality local journalism. Become a member of Crosscut today!

Comments:

Posted Tue, Oct 19, 1:29 p.m. Inappropriate

Also pollsters discount things like this:

http://www.youtube.com/watch?v=SeOHPfsCtFo

Which get passed around the internet, so that even women who might not bother to vote get motivated. How that impacts either party is still up for grabs, but historically women tend to vote Democratic.

GaryP

Posted Tue, Oct 19, 2:48 p.m. Inappropriate

The Fabrizio poll you referenced was in the field nearly 3 weeks ago. The Moore poll results mention in reports today is also old data. The Rasmusssen poll from 9/28 showed a Rossi lead. Rasmussen now shows a 3 point Murray advantage marking a 6 point increase in Murray's support since 10/6. The PPP poll released today shows a 2 point Murray advantage. Both Rasmussen and PPP assume a modest increase in the percentage of Washington voters expressing a Republican prefernce.

One interesting fact buried deeper in the PPP results--among those polled who say they have already voted Murray commands a 52 to 47 lead over Rossi.

Information filtering out from California shows that Democrats are turning out for early voting in greater numbers than Republicans. AP reports today show that the Democratic turnout is higher than at this point in 2006 causing speculation that Democratic turnout in that state will rival 2008.

If Democratic turnout in Washington and Oregon matches California, there may be no Republican wave on the West Coast.

quiller

Posted Tue, Oct 19, 3:10 p.m. Inappropriate


A couple of years ago I was watching PBS's Washington Week. It was a few days following the New Hampshire primary, and the first topic for discussion was how wrong the polls were (Clinton defeated Obama). One panelist described how she had already written most of her story for the next day's edition, but had to burn the midnight oil to completely rewrite it. All the reporters sitting around the table shook their heads and expressed a combination of guilt and chagrin over their faith in the polls.

Ten minutes later in the program, the same panelists were saying things such as, "The current polls on white males between the ages of 25 and 40 are saying that . . ."

I was astounded. The major polls all got New Hampshire wrong, and those polls were your basic "Who will you vote for?" variety. But it didn't matter, the members of the Washington press corp who are always featured on that program just went on talking about polls allegedly describing the thinking of very limited population segments. The opening story did not make a dent in their thinking, they just went on their merry way expressing so much confidence in the new information they were sharing. I was disheartened by their lack of anything original to say beyond what the pollsters were feeding them.

I have three words for news outlets that insist on reporting poll numbers instead of performing analyses of issues, previous voting records, or past business connections: Dewey Defeats Truman.

Lindy

Posted Tue, Oct 19, 3:25 p.m. Inappropriate

Quiller-

The Moore numbers are not "out of date."

The PPP poll's sample is only +3 for the Ds and it shows the race within the margin of error - basically a dead heat. I think this is credible, far more credible than the Elway or CNN poll.

Posted Tue, Oct 19, 3:46 p.m. Inappropriate

In reference to Lindy's piece, what disappoints me is that polls are considered "news" and subsequently used to support a narrative. It's lazy, stupid, jounalism. There is no longer any analysis of candidates positions, and much more importantly, the candidates understanding of those positions.

Bella

Posted Tue, Oct 19, 4:27 p.m. Inappropriate

I'm not seeing wildly inconsistent polls when it comes to what matters: who is winning.

Every recent poll has Murray winning: 49-51%. In other words, they are all saying pretty much the same thing, when you consider the margin of error.
Elway, CNN, SurveyUSA, Washington Poll, and even Rasmussen.

The Fabrizio poll Vance references is a joke within this context. It shows Rossi leading only after slinging a series of enormous stinking mudballs at Murray.

Rossi is slinging them, for sure. But Murray seems to be hitting back effectively.

Rossi's negatives appear to be way too high for someone who has been out of government for so long.

Jan

Posted Tue, Oct 19, 4:32 p.m. Inappropriate

Chris, I presume you understand the difference between a likely-voter poll and registered-voter poll, however you don't distingish the two in advising us of your reliability of one poll over another poll. In particular you don't think CNN is credible, however you don't cite which CNN poll is the non-credible poll. Is it the CNN likely-voter poll or the CNN registered-voter which you consider not to be credible? Is it both?

Bella

Posted Tue, Oct 19, 8:14 p.m. Inappropriate

Jan: Most of the polls have Murray "ahead," but within the margin or error and under 50%. That is far different than the Elway and CNN polls.

Bella: I am referring to the CNN/Time poll of likely voters that showed Murray ahead by 8%. I am not aware of any CNN poll of all registered voters in this race.

Posted Tue, Oct 19, 9:32 p.m. Inappropriate

Chris, thanks for the reply. The CNN/Time/Opinion Research Corp. (full name of the polling company) is one of the few pollsters that concurrently conduct both likely-voter AND registered-voter polls. The 08-12 Oct. CNN likely-voter poll you referenced has Murray with an 8% lead, however the results of the CNN registered-voter poll undertaken during the same time period (08-12 Oct) as that CNN likely-voter poll shows Murray with only a 4% lead. The sample size of the registered-voter poll was 1340.You should be easily able to research and confirm this.

I am surprised you appear completely unaware that CNN produces a likely-voter poll AND and a registered-voter poll.

Bella

Posted Wed, Oct 20, 6:11 a.m. Inappropriate

Bella:

Its the same poll. What gets reported is the result among likely voters, and what I am questioning is CNN's definition of likely voters.

Posted Wed, Oct 20, 7:50 a.m. Inappropriate


Chris, you are incorrect. There is a difference between likely voter polls and registerd voter polls. It is NOT the same poll. If you look, the sample size for the likely-voter poll was 850. I informed you in my reply above the sample size for the registered-voter poll of the same period was 1340. The only thing the same are the dates when the polls were conducted and reported. The results from CNN's likely-voter poll AND and their registered-voter are REGULARLY reported. I suspect the fact you were unaware that CNN conducts both a likely-voter poll and registered-voter poll might be due to where you are regularly sourcing(RCP?)your poll results. I might suggest going to the actual polling company site or find a more comprehensive poll result reporting site than the one you are currently using.

Bella

Posted Wed, Oct 20, 12:06 p.m. Inappropriate

Bella,

I looked directly at the information CNN released. I think you and I are just describing the same thing differently. CNN polled 1340 registered voters. Of those they believe 850 are likely voters. The results among LVs are what get reported in the media and on the internet. And I think their sample of LVs includes too many Democrats.

Posted Wed, Oct 20, 1:04 p.m. Inappropriate

Chris, I still think you are incorrect, however please do us a favor. Please paste the link to the CNN poll results you found so we can all
take a look at it.

Bella

Posted Wed, Oct 20, 2:05 p.m. Inappropriate

http://i2.cdn.turner.com/cnn/2010/images/10/13/topstate6.pdf

There you go.

Posted Thu, Oct 21, 7:45 a.m. Inappropriate

Chris, thanks for the link to the results. We are able to confirm there are posted results from the sample of the polling of registered voters and likely voters. The results are different and the link shows the reported results.
The LV's reported results are NOT the only results that are reported in the media and internet. The results of the RV's and LV's are available to the media and internet outlets. It is up to the outlet to decide if they want to include them in their reporting. I suspect those reporting sites that are engaged in
serious analysis of electoral polling will include them. Those whose interest in reporting the horserace or creating a narrative will probably not include the results. Lastly, I do understand that all the polling companies I am aware of eventually change their polling from registered voter to likely voter the closer we get to an election.

Bella

Posted Thu, Oct 21, 10:31 a.m. Inappropriate

I guess Rush Limbaugh wasn't available to provide is unbiased assessment of this race so you had to turn to Chris Vance,chairman of the state Republican Party for his unvarnished view on the matter.

What a hoot.

Login or register to add your voice to the conversation.

Join Crosscut now!
Subscribe to our Newsletter

Follow Us »