Saturday, September 27, 2008

Another Internal: Walberg 50, Schauer 40

From the Battle Creek Enquirer:
Tim Walberg has his own poll, and according to that one, he's leading the 7th Congressional District race.

The campaign for the Tipton Republican released the results Thursday of a Sept. 15-16 poll, only hours after the release of polling information from state Sen. Mark Schauer, D-Bedford Township, Walberg's main opponent.
In addition to the Enquirer, Chetly Zarko got the memo from the Walberg campaign.

It was conducted over September 15 and 16 by National Research, Inc., a Republican polling firm. It has a sample of 300 "likely voters" stratified geographically, and has a margin of error of 5.7 percent.

Here are the results (with 7/08-09 results in parentheses where available, also +/- 5.7):

Mark Schauer vs. Tim Walberg

Mark Schauer (D) - 40 (31)
Tim Walberg (R-inc.) - 50 (47)

Walberg Favorable - Unfavorable

Favorable - 57
Unfavorable - 29

Schauer Favorable - Unfavorable

Favorable - 33
Unfavorable - 18

Obama vs. McCain

Barack Obama (D) - 40
John McCain (R) - 55

Generic Congressional Ballot

Republican - 46
Democrat - 41

So, what does all of this mean?

First, it means that either A.) the last Walberg internal was an outlier or B.) Mark Schauer has made an enormous nine-point gain since the beginning of August. The first one isn't a big deal-- it made Walberg's people feel good about themselves for a day. The second one should worry them a lot. Though, then again, here's what Walberg's pollster says:
Despite the barrage of negative ads directed against him, Congressman Tim Walberg leads Mark Schauer 50%-40% according to our most recent congressional survey in the district, conducted on September 15 and 16 among 300 likely voters in the district. This represents a significant gain for Congressman Walberg, who was garnering 47% in our July survey. With his poll numbers now at 50%, Walberg enters the month of October with momentum.
(Emphasis added.)

That's quite a positive spin! Personally, I'd say that the significant gain goes to the one who, you know, actually gained more.

It's also important to note that, with a 10-point lead in a poll with a margin of error of 5.7 percent, Walberg is now within the margin of error (barely) in his own polling. That's not what will get the headline, though. The magical 50 percent mark is an important one to the media, and one that I don't think Walberg has reached in any previous polling. (He also didn't reach it in 2006-- he had 49.99 percent of the vote).

Last night, I wrote a partial defense of internal polling, saying that it's still useful and reliable. Now, I'm wondering if there might actually be a systematic difference between Schauer's polls and Walberg's polls-- namely, in the geographic breakdown and in the likely voter screen. Is Walberg oversampling Branch County and undersampling the youth vote? Or is Schauer oversampling Calhoun County and undersampling Hillsdale County? I don't know, and the campaigns don't release that kind of information.

The reason I wonder about that is the name recognition. Only 55 percent know who Mark Schauer is in this poll, where 67 percent did in Schauer's last poll. Similarly, 86 percent know who Tim Walberg is in this poll, where it was only 76 in Schauer's poll. That kind of difference surprises me, and makes me think someone weighted the counties differently.

Then again, both polls strike me as plausible, especially with the large margins of error. It really doesn't matter, though. There are 38 days until election day. That's more than enough time for everything to change.

Overall, it was a good poll for Walberg, but maybe a better poll for Schauer.

Labels: , , ,

The Walberg poll was conducted by a respected independent firm, which I would put a little more faith in then the left-wing propoganda organization that conducted Schauer's push poll. When a polling firms claims it exists to elect Democrats and advance liberal ideas it ceases to be reputable. Walberg's firm may work with Republicans but they also work with Fortune 500 companies and seem to provide a scientific poll. No telling what kind of "background" the Myers pollsters read to participants to get the results they were looking for.
Well, no. As I've said before, it's not in the polling firm's interests to give rigged data. That's just plain stupid, and Democrats, believe it or not, aren't stupid. Bad data does nothing for a campaign except get a couple of good newspaper articles no one reads. Good data lets a campaign know if their strategy is working. If Myers has been supplying bad information, then he'll soon find himself with fewer clients.

Likely voter models and geographic distributions are probably what account for the differences, plus just random chance. Statistically speaking, both of these polls are plausible. And even if they're wrong, that just means they're the 1 in 20 that are naturally wrong. That's how statistics works.

Neither poll said anything about reading "background" information, so I'm inclined to take them at their word that they just asked a straight Walberg-Schauer matchup question. The EPIC-MRA polls do provide biographies, but those are provided after the first matchup question, and the results post-biographies are reported separately.

I'm trying to be reasonable here. I think both polls probably have some merit, though I'd appreciate it if the campaigns would be a little more transparent and provide more data. But both Democratic and Republican firms can provide real data, and often do. Any campaign paying for fake data is really wasting its money.
Post a Comment

Subscribe to Post Comments [Atom]

<< Home


August 2006   September 2006   October 2006   November 2006   December 2006   January 2007   February 2007   March 2007   April 2007   May 2007   June 2007   July 2007   August 2007   September 2007   October 2007   November 2007   December 2007   January 2008   February 2008   March 2008   April 2008   May 2008   June 2008   July 2008   August 2008   September 2008   October 2008   November 2008