Wednesday, May 16, 2012

Garbage Marquette poll alert! It even contradicts itself.

Seems like I gotta do this every day now. I had a feeling Charles Franklin just might adjust things in Scott Walker's favor in his latest Marquette University poll, and he sure did. Franklin's latest poll says Walker leads Barrett 50-44, and by an even higher 51-43 among registered voters. And it doesn't take very long to look at the crosstabs to see exactly why.

The first tip-off is the same as yesterday's faulty PPP poll- Party ID and ideology.

Marquette Law poll, May 2012
Party ID
Republican 28
Democrat 29
Independent 40

Independents' "Leanings"
Republican 41.4
Democrat 36.0
Neither/Indpendent 21.5

Ideology (among those answering)
Conservative 49
Moderate 31
Liberal 20

49-31-20 is a whole lot different than the 2010 exit poll breakdown of 36 con- 43 mod- 21 liberal. And it also tells you that a lot of those alleged "Independents" are really Tea Baggers who refuse to admit they're Republican. An experienced pollster like Franklin should be seeing that poll's right-wing leanings and adjust poll respondents accordingly, but he chose not to.

So let's use that 2010 exit poll and see what the results change to.

Walker 46.52, Barrett 45.66, virtual toss-up.

Pretty different scenario, isn't it? So why the difference? Well, Franklin insists that Walker supporters are more fired up to vote.
“While both parties show unusual levels of involvement in the campaign, Republicans appear to hold an advantage in likely turnout, although Democrats are more likely to have been contacted by a campaign. In a close election with so few undecided voters, enthusiasm, turnout and campaign contact with voters may make the difference.”
Except that Franklin's own numbers don't say that when you go into ideology.

Do you plan to vote in recall election?
Very conservative- 88.2% absolutely, 5.2% very likely
Conservative- 82.5% absolutely, 9.0% very likely
Moderate- 88.6% absolutely, 6.6% very likely
Liberals- 82.5% absolutely, 4.9% very likely
Very liberal- 89.6% absolutely, 7.0% very likely

Seems like all classes of voters are planning to get out to the polls, so why make the poll sample heavy on conservatives? Again, Franklin made a choice to do that.

The poll also has other demographics that are skewed and favor Walker.

Age of respondents
18-29 8%
30-44 13%
45-59 33%
60+ 45%

In 2010, the 18-29 vote was 15% of the total, and more than 1/3 of the voters were under 45 years old, instead of the 1/5 that Franklin thinks they'll be. Doubly hilarious is Franklin's assertion that a majority of those under 30 will vote for WALKER. The only people under 30 who support Walker are the no-talent hacks that work for him and other right-wing political organizations. No real person under 30's voting for that guy. Pathetic.

Who did you vote for in 2010 election
Walker 51.6%
Barrett 40.8% (actual result, Walker 52.25-46.48)

Franklin assumes that nearly 90% of the people who will be replaced in the recall election are Barrett voters. That is an absurd assessment on its face, because if you voted for the other guy, I'm guessing you might be a bit pissed off at Scott Walker these days, and I don't think you'll be the one who bails on the election. Amazingly, Franklin has no cross-tabs on this question, and left off a question that he had last time asking how many respondents had signed a recall petition. Both of these things make the poll seem even more fishy.

And the last one that's worth mentioning is the geography of the poll respondents, which shows another pro-Walker Franklin skew.

What region are you from in Wisconsin?
City of Milwaukee 12%
Rest of Milwaukee media market 30%
Madison media market 17%
Green Bay/ Appleton media market 20%
Rest of state 21%

Franklin also deflated that Madison media market number from an actual of over 18% of respondents. Well, if nearly twice the amount of voters come from the 262 vs. the Madison area, well yeah, Walker will do well. But does anyone honestly think that will happen, especially when past data shows that there are more voters in Dane County than Waukesha County? Then again, the fire that we have here in the Madtown area has been underestimated from Day 1, so why would this be any different?

What's discouraging about all this is that there's no pushback, and no replying polls that use actual turnout data, so the brain-dead media runs along with these figures and is now going to try to convince people "Walker's ahead." It's frustrating when you're a person who lives off of numbers and raw data, but instead find an administration and a right-wing propaganda machine that is allowed to try to create their own reality.

The goal with both these flawed polls and Walker cooking the job books is to trick the average bystander, and drive down our pro-recall enthusiasm. They're doing this because they know, behind the misleading topline results, the crosstabs show that if the good guys show up at the polls in 20 days, Tom Barrett will be elected Governor. We need to stay strong, fight back, and make this happen.


  1. Well, this is one post where I'm going to have to strongly disagree with something you wrote.

    Most of the people I know who are under 30 and voting for Walker aren't "no-talent hacks that work for him and other right-wing political organizations."

    They're blue collar white people who didn't go to college.

    Of course, blue collar white people of all ages are trending against Democrats. Doesn't mean that the Democrats shouldn't do a better job trying to connect with them.

  2. Alex- Maybe I'm a bit over the top there. You are correct that Dems should be doing serious outreach to young voters (who went 2-1 for Obama).

    How do you do that? By pointing out th absurdity of voter ID, by actively backing marriage equality (which Tom does), and telling young blue-collar whites just how important unionization is in terms of getting them a decent wage. Because we all know right-to-work-for-less is coming if this state is dumb enough to re-elect Walker.