OK

If you have perused our awesome new Poll Explorer feature, you already know that the last few days have brought some examples of better-than-average polling news for the Democrats.

For the first time in a very long time, the Democratic nominee has been shown in the lead in polls in gubernatorial races in Arkansas and Michigan. Marquette confirmed their earlier poll showing that the Democrats have a very real shot at knocking off Gov. Scott Walker in Wisconsin. Republicans also had to be a bit surprised to see that Rasmussen gave Democratic Sen. Mark Pryor the slightest of leads over Tom Cotton in Arkansas, a race that most GOPers have been pretty confident about.

But maybe the banner headline was a double whammy of data that came out of the heartland. In our inaugural Polling Wrap of the cycle on Tuesday, we talked about how the release of internal polls can often drop us hints about the relative strengths of the parties heading into an election. In the case of Kansas Gov. Sam Brownback, however, his internal poll dump, coming in the immediate wake of a SUSA poll showing him losing to Democrat Paul Davis by a fair margin, raised more questions than it answered.

Before we look at Brownback's plight, however, let's grab the numbers, in the form of all the general election polls released to the public in the last three days (Aug 26-28):

AR-Sen (Rasmussen): Sen. Mark Pryor (D) 44, Tom Cotton (R) 43

IA-Sen (PPP): Bruce Braley (D) 41, Joni Ernst (R) 40

IA-Sen (Suffolk): Bruce Braley (D) 40, Joni Ernst (R) 40

KS-Sen (SurveyUSA): Sen. Pat Roberts (R) 37, Chad Taylor (D) 32, Greg Orman (I) 20

ME-Sen (PPP for Shenna Bellows): Sen. Susan Collins (R) 57, Shenna Bellows (D) 33

MI-Sen (EPIC-MRA): Gary Peters (D) 45, Terri Land (R) 39

MI-Sen (Mitchell Research): Gary Peters (D) 46, Terri Land (R) 44


AK-Gov (Rasmussen): Gov. Sean Parnell (R) 47, Byron Mallott (D) 36

AZ-Gov (PPP for Progress Now Arizona): Doug Ducey (R) 35, Fred DuVal (D) 35, Barry Hess (Lib) 12

AR-Gov (Rasmussen): Mike Ross (D) 46, Asa Hutchinson (R) 44

IA-Gov (PPP): Gov. Terry Branstad (R) 48, Jack Hatch (D) 35

IA-Gov (Suffolk): Gov. Terry Branstad (R) 47, Jack Hatch (D) 35

KS-Gov (Cole Hargrave Snodgrass for Brownback): Gov. Sam Brownback (R) 43, Paul Davis (D) 42

KS-Gov (SurveyUSA): Paul Davis (D) 48, Gov. Sam Brownback (R) 40, Keen Umbehr (Lib) 5

MA-Gov (Social Sphere): Charlie Baker (R) 38, Martha Coakley (D) 37

MA-Gov (Social Sphere): Charlie Baker (R) 37, Steve Grossman (D) 33

MA-Gov (Social Sphere): Charlie Baker (R) 44, Don Berwick (D) 19

MI-Gov (EPIC-MRA): Mark Schauer (D) 45, Gov. Rick Snyder (R) 43

MI-Gov (Mitchell Research): Gov. Rick Snyder (R) 47, Mark Schauer (D) 46

PA-Gov (Franklin & Marshall): Tom Wolf (D) 49, Gov. Tom Corbett (R) 24

SC-Gov (Rasmussen): Gov. Nikki Haley (R) 51, Vincent Sheheen (D) 36

WI-Gov (Marquette): Mary Burke (D) 49, Gov. Scott Walker (D) 47


No House polling was conducted in the past several days.

A few thoughts, as always, await you just past the jump ...

Releasing internal data in the wake of a bad public poll is pretty old hat in the campaign game. It is an attempt, transparent though it might be, to change the narrative after the public gets wind of a survey that makes your candidate's prospects look bleak. So, the fact that Team Brownback hit the send button on their recent internal poll in the immediate wake of the SUSA poll showing them down eight points to Democrat Paul Davis was not an unprecedented move.

But it also might not have been a great move, either.

For one thing, let's look at the topline: Brownback 43, Davis 42. It is not exactly a supreme sign of confidence to release a poll making the following essential argument:

That poll showing us losing?! Total bullcrap, people. Look at THIS poll. We are ahead! By ... um ... a single point. Victory is around the corner!"
Ultimately, one has to wonder if Brownback was better off sitting on that particular poll. A one-point lead in his own polling did absolutely nothing to quell the notion that he was in deep, deep trouble.

Another reason why this wasn't the swiftest of moves was the polling memo itself, which was wholly unconvincing. Did you know that Davis is little known? And that Obama is unloved in Kansas? Even if both of those things are true (and there is no reason to believe that they are not), an argument can be made that this actually makes Brownback look worse. Another way to frame that:

So, Governor Brownback, you are the Republican governor of a red state where President Obama is pretty much loathed. And your own polling shows you only up by a single point over a guy you are basically calling out as a Some Dude?! How much must you be reviled for that to be the case?!
What's more: this poll had a defining characteristic you don't often see in the game. It was weighted by political party. All polls with weighting are based, at least in small part, on the assumptions of pollsters. For the last two cycles, for example, Gallup messed the bed by assuming in their polls that the electorate was going to be a good deal more white than the actual electorate turned out to be.

But weighting by partisan identification adds an entirely new level of potential mischief, because other factors upon which pollsters typically weight polls (gender, race, age) are not nearly as fluid as political party identification. Republicans learned that the hard way in 2012, when the unskewing phenomenon was based on a nearly uniform belief that pollsters were erring by assuming a greater proportion of Democrats would turn out at the polls than Republicans. As it turned out, the pollsters were right and the "unskewers" were wrong. A lot of Republicans had redefined themselves as Independents, which led to Romney doing better with Independents than McCain, but still losing because only 32 percent of the electorate thought of themselves as Republicans.

What assumptions did Brownback's pollsters make regarding the Kansas electorate? Perhaps they modeled it after 2010, but that's awfully optimistic—it's going to skew things (pardon the use of that particular word) by assuming the electorate will mimic the best Republican electoral cycle in generations.

One thing is for certain: they were highly unlikely to have been bullish on Democratic participation. Which is just one more reason why that poll actually made it easier to assume that Brownback is in genuine peril in ruby-red Kansas, rather than alleviating any Republican fears about his potential ouster.

Originally posted to Daily Kos Elections on Fri Aug 29, 2014 at 05:30 AM PDT.

Also republished by Daily Kos.

EMAIL TO A FRIEND X
Your Email has been sent.