The new Marquette Law School Poll run by Charles Franklin was released yesterday (Wednesday, September 19). It didn’t take Thompson and the GOP long to attempt to discredit the poll – one they have always held in high esteem. The reason – the new poll told them something they didn’t want to hear.
On August 12, 2012 in an interview on Capitol City Sunday (Madison WKOW-TV), Thompson had this to say about the Marquette Poll:
“I think the Marquette poll which is considered pretty much the golden rule or the gold standard, came out and says I was 8 points up. So, I feel very good about going in, and I sincerely believe the momentum is with me and I feel very comfortable about the election.”
On Tuesday, the day before the poll was released, Thompson expressed confidence:
“Let’s see what the Marquette poll says tomorrow,” he said. “I feel very good about where I am at – feel very comfortable about the future of this race.” (MJS article)
After the poll was released, Thompson spokeswoman Lisa Boothe accused the poll of “oversampling for Democrats,” giving an edge to Baldwin. From the Thompson statement:
“We do not believe that today’s Marquette poll reflects the opinions of Wisconsin voters,”
An analysis of the poll shows in this most recent sample, there was a slightly higher number of respondents who identified as Democratic. While Franklin adjusts slightly for that in post-stratification, there is still a slightly higher number identifying as Democrats. The Thompson/GOP vilification is unwarranted. The Marquette poll over the past four months has been very consistent, showing an emerging pattern in the electorate conservatives will be desperate to stop. To the tables…
Marquette Law Poll Demographics (link to graph)
The demographics of the poll remain extremely consistent from July-September. The 45+ age group is slightly over-represented, as is respondents with a post-secondary degree. In political party identification, it is the norm in the poll for Republicans to be slightly over-represented; while Independents are under-represented. The September poll is an exception – showing Dems slightly over-represented. While this is leading to Thompson dismissing the poll – he should not. Even as the party ID appears to lean more Democratic, the big picture of the poll shows there is more at work in the results.
Marquette Law Poll Political trend (link to graph)
The poll shows consistently that about 66% of respondents (low 63-high 69%) follow politics “most of the time,” and 83% are “absolutely certain” to vote in the November election (low 81-high 85%). The next set of numbers disputes the Thompson/GOP bias claim in the poll – political leaning.
The fluctuation of respondents identifying as Conservative, Moderate, or Liberal is relatively consistent in a poll with a +/-3.8% margin of error (MoE). From the August 16-19 poll to the September poll, the values of respondents identifying is within the MoE for each category respectively – Conservative (41-39%), Moderate (30-32%), and Liberal (23-26%). The shift from Thompson to Baldwin has been with moderates, and also with those who previously “did not know enough” about Baldwin, now changing to voting for her – and against Thompson. The slight increase in “Liberal” or “Democratic” respondents does not account for the huge swing to Baldwin.
Marquette Law Poll Favorable rating (link to graph)
The favorability rating of the candidates and Scott Walker show part of the story. Barack Obama’s favorable rating has increased from 51-55%, Baldwin’s from 26-36%. Mitt Romney’s favorable rating is flat at 36%; while at the same time Scott Walker’s has fallen from 50-46%, Tommy Thompson’s from 44-39%.
Marquette Law Poll Unfavorable rating (link to graph)
The unfavorable rating is telling as well – showing a consistent shift apart from any alleged Democratic bias in the sample from a single poll, at a relatively insignificant level. Barack Obama’s unfavorable rating has gone from 42% down to 39%, Baldwin’s has stayed steady at 31%. Romney’s has increased from 42% to 51%, Walker from 45-49%, and Thompson from 36-44%. All increases in Republican unfavorable ratings are outside the MoE – and shifts in “conservative” or “liberal” political ID.
Marquette Law Poll Pres/Senate likely voters (link to graph)
Amongst likely voters, there has been a dramatic shift in the Presidential and Senate campaigns. Much of that shift can be attributed to the GOP Convention disaster, Romney’s campaign failures, and the Dems convention and messaging success. Romney/Ryan received a slight bump, closing them to within 49-46% mid-August. The recent poll has Obama/Biden up 54-40% . The same bump had Thompson up 50-41% on Baldwin; the recent poll has a huge swing, putting Baldwin up 50-41% on Thompson. This eighteen-point swing simply is not explained away by the nominal oversampling of Democratic voters. There is not enough shift in “liberal” vs. “conservative” to account for such a change.
Marquette Law Poll Issues (link to graph)
The August 2-5 poll showed 43% of respondents would agree to “higher taxes for more services.” This recent September poll shows 47% agreeing to “higher taxes for more services.”
In the same August poll, 44% approved of “Obamacare,” in September the number had risen to 49%.
The Bush tax cuts popularity continued decline, from 35% in August to 30% in September.
The above surveyed issues in the poll demonstrate a further shift consistent with the data – the electorate in Wisconsin is trending away from the right-wing GOP issues and candidates. That is why they attempt to discredit the Marquette Poll. The pattern playing out with conservatives is becoming clear.
When facts, data, and science don’t support your ideology-driven agenda, there are three courses of action as a right-wing conservative corporatist. First, ignore. Second, pay someone to write a study that supports your agenda. Third, pay to have your lies repeated to nationwide media.
The full polls cited above can be found here:
Help keep independent, progressive journalism alive!