Just read up on the newest Marquette University poll which shows a tight governor's race at this time, with Scott Walker leading Tom Barrett by 2, Kathy Falk by 4, Doug La Follette by 7, and Kathleen Vinehout by 8 (interestingly, Walker is under 50% in all races). This is a closer score than we had with Marquette's last poll, which had Walker up at least 6-7 points on all comers, which caused me to shred Charles Franklin's poll for oversampling right-wingers and old voters.
Well, much like the poll results, Franklin's gotten rid of some of the excesses, but he's still oversampling Scott Walker's voters. And as usual, I'll break down how, and I do want to thank Franklin and Marquette for being transparent enough to show their methodology.
Ther first tip-off comes from the Topline demographics, and I'll break down 3 in particular: gender, ideology, and age.
First of all, the gender breakdown in the poll is 52% men, 48% women, and strangely Franklin gave men a higher weighting, as the actual respondents were 51-49% men. This goes against the largely accurate 2010 CBS News Exit poll, which had the split at 50-50. On the surface, a 2% difference doesn't seem like much, except that you look further at the poll and realize Walker's approval is +12 among men, and -6 among women, and Falk's favorability is -20 among men but only -3 among women. This gender gap is quite obvious in the heads-up races:
polls by gender
Walker vs. Falk men: Walker +12.4, women: Falk +4.4
Walker vs. Barrett men: Walker +8.3, women: Barrett +6.0
Walker vs. Vinehout men: Walker +16.2, women: Vinehout +0.8%
Walker vs. La Follette men: Walker +15.3, women: La Follette: +1.0
Guess that 14-16 point gender gap tells you that Walker is in a world of hurt if women use the superior numbers in the population to get out and vote, doesn't it?
And the leaning towards men is evident in the skewing of some of the other numbers. The higher amount of pro-Republican men also seems to help explain the higher amount of conservatives being part of the poll. Among those answering the question, the ideology breakdown is as follows: 20% liberal, 39% moderate, 41% conservative. Now this is better than January's debacle, which was 22-34-44 liberal-moderate-conservative, but it still doesn't come close to matching the 21-43-36 voters in the 2010 elections.
If there are more moderate voters, it massively hurts Scott Walker's poll standing, as Walker's approval rating is 43% among moderates, and he loses moderates by nearly 18 to Barrett, 11 to Falk, 7 to Vinehout, and 6 to La Follette. In fact, if you use the 21-43-36 parameter in the 2010 exit poll, the Falk and Barrett races become quite interesting indeed.
poll using 2010 exit poll weighting
Falk 46.8 - Walker 46.6
Barrett 47.6 - Walker 44.9
Walker approval- 48.7 - 48.6 (vs. 50.4-46.6)
Think the news stories on the polls might be a bit more interesting if Walker was losing by 3 to Barrett, tied with Falk, and had his approval-disapproval dead even? Again, it's not an egregious error by Franklin and Marquette, but a 4-5 point swing is a lot in the race's perception right now.
Lastly, the Marquette poll only has 7 percent of voters under 30, and nearly 60% of voters at 45 or older, and that's after deflating the actual respondents, of whom 77% (!) were 45+. So let's compare that to the GAB's list of registered voters after the 2010 elections, and the 2010 CBS exit poll.
Voters under 30 and over 45, estimates
Marquette under 30 - 6.9%
GAB under 30 - 9.5%
CBS 2010 exit poll- 15%
Marquette over 45 - 59.6%
GAB over 45 - 57.2%
CBS exit poll - 62%
While the older vote seems to be in the ballpark, the younger vote seems to be underestimated in the Marquette poll, and that's no small deal, because 65% of those polled under 30 years old DISAPPROVE of Walker, Barrett beats Walker 3 to 1 among young voters, Falk beats Scotty by nearly 40, and Vinehout and La Follette beat Walker by well over 30. Needless to say, if young voters get out and vote above their numbers, Walker will have a very tough time winning, but the Marquette poll underplays this huge deficit Walker has among the young, because young people simply don't answer phone calls from weird numbers.
So when you look at the Marquette poll, they've clearly made some tweaks that don't make the numbers seem as absurd as they did 2 months ago, and they get minor credit for that. But they're still overrepresenting conservatives and underrepresenting moderates and young voters - two constituencies that don't like Scott Walker. So when you look at those poll numbers, it may be fair to knock down Walker another 2-3 points, and when you realize that most Wisconsinites have made up their mind when it comes to Scott Walker, that 3 points may put him in a major hole that he can't get out of.
The challenge for the Dems is now to make sure they maintain the fire, and get young voters and women out to the polls on June 5. If that happens, Wisconsin will be likely to make history by being the third state to kick out its governor through a recall. Now can the Dems continue to act in a way that makes people want to get out to the polls and vote for them, or will they be a top-down, backroom-deal group that doesn't understand how the game has changed in the last year? We'll see.
No comments:
Post a Comment