What are my thoughts on the CNN poll that showed Jacky Rosen leading Dean Heller, 47-43, and Steve Sisolak ahead of Adam Laxalt, 45-41?
So glad you asked.
The velocity with which Team Laxalt took to Twitter to knock down the survey was interesting, to say the least. Pollster Chris Wilson, whom I respect, is saying that the likely voter sample can’t be right because it is too high a percentage for a midterm election.
First, he may be right. I am not sure I like the culling of voters method used by this pollster, which is different than many others because the firm called people who were not registered, too. But that also does not mean the poll is wrong. The pollster found almost 700 voters likely to vote, which is a large statewide sample (4.6 percent margin of error) if it is accurate.
Second, Wilson is right that a “+6 Dem” sample also probably wouldn’t be optimal to obtain the best numbers – voter registration is +5 Dem, to use his shorthand, and probably will be less on Election Day because GOP turnout is almost always larger in a midterm.
But, alas, we don’t know that’s what the sample is because the actual numbers are 39-34, Dem, in Nevada — and the sample has 42 percent of respondents not identifying themselves as being with a major party (the actual number is about 27 percent). That likely means some registered Democrats and some registered Republicans were not telling the truth about their registration, so we really don’t know what the partisan breakdown is.
So having addressed those criticisms – and I would still like to know the actual partisan breakdown — let’s be clear about something: The poll validates other data out there, and I wouldn’t be surprised if it reflects Wilson’s own numbers. The leads are within the poll’s margin of error, just as all other polls have been.
Both races are close; this we know.
Even if you take Wilson’s criticisms at face value, the registered voter numbers also are not good for Republicans – just a 3-point lead for the Democratic candidates instead of four.
Whether this model is closest to what will occur on Election Day, or Wilson’s is, or other pollsters’ are, this survey is not an obvious outlier. Does this prove the races are starting to break the Democrats’ way, just a little?
Not necessarily. But any experienced hand out there in this environment and knowing history realizes this is approximately the time – less than three weeks before early voting starts – that races can and often do begin to break one way or another.
Some other thoughts:
—SSRS, which conducted the survey, is not a traditional political polling operation, as you can see by its client list. But that doesn’t mean the company doesn’t adhere to accepted industry practices; it just means campaign polling is not what it is known for. In fact, the renowned 538 site rates the outfit an A-minus with a very slight GOP bias.
—The approval ratings for the Senate candidates are very much in line with other private and public polling:
Rosen is barely above water (43-38) and Heller is barely under (43-46).
President Donald Trump (45-51) is where he has been for awhile.
—The gender gaps are huge. Considering the backdrop of #MeToo and the Kavanaugh hearings, if this poll is accurate, Laxalt and Heller have a lot to worry about. Here’s why:
Men: 48-41, Heller; 44-41, Laxalt
Women: 52-38, Rosen; 50-38, Sisolak
—I have mentioned in the past that Heller and Laxalt hope to drive up the white vote. Here’s why:
Whites, 49-41, Heller; 48-42, Laxalt.
—CNN leaving independent contender Ryan Bundy out of the governor’s race matchup may not be an insignificant error. He has been at 4-6 percent in most polls I have seen or know about. Of course, that could diminish by Election Day. But he’s a factor — small, medium or large — that is not measured here. (By the way, did you know Bundy has erected campaign signs in Las Vegas?) He will take votes from Laxalt, so it’s possible this poll is even better for Sisolak, but hard to tell.
—Health care (24 percent) is the most important issue in the Senate race by a nose over the economy (22 percent). Immigration is third (18 percent). If health care stays at the top or moves higher, it could be sayonara for Heller.
I wish I had more information for the poll— regional and minority breakdowns, for instance. But this is a live-caller poll conducted by an experienced market research firm, so any knee-jerk dismissals should be taken with a big grain of salt.
Disclosure: Steve Sisolak has donated to The Indy. You can view a full list of our donors here.