New Marquette Poll Has Scott Walker Up By 7 Points Over Tom Barrett

**Updated 7 pm.

If Gov. Scott Walker can keep up his momentum going into , the numbers indicate he will come out the winner over Milwaukee Mayor Tom Barrett.

A Marquette University Law School poll released Wednesday shows Walker leading by 7 points over Barrett - 52 percent to 45 percent - among likely voters. Two weeks ago, Walker led Barrett 50 percent to 44 percent .

Most of the 720 registered voters were interviewed before Friday’s first gubernatorial debate using landlines and cell phones and the results are taken from the answers of 600 likely voters. The margin of error is plus or minus 4.1 percentage points.

Another poll cited in the Washington Post is from Celinda Lake and was commissioned by the Greater Wisconsin Committee's political fund. Using results from an automated phone survey of 600 likely voters between May 24 and May 28, the numbers show Walker and Barrett in a tie at 49 percent each.

Marquette professor Charles Franklin cited Lake's poll as he compared other poll results to that of the Marquette Law School poll since January.

"Up until two weeks ago, we had a flurry of polls," said Franklin, who oversaw Marquette's poll. "What's interesting to note is that Democratic polls, which are conducted by professional and respected pollsters, shows a tighter race, but not with their guy ahead."

The Marquette was released during edition of "On the Issues" with Franklin and WISN-TV's Mike Gousha.

Gousha asked Franklin to talk about jobs and how voters view the state's jobs situation.

"In the last two weeks, both campaigns are battling over jobs numbers," Gousha said. "Both sides claim their numbers are correct, so what did we find this time with jobs perception amongst respondents?"

Franklin noted that current events can drive the numbers. Two weeks ago, 20 percent said the state gained jobs; 38 percent said no change and 37 percent said the state lost jobs.

But since the new sets of numbers from both the Bureau of Labor Statistics and the quarterly figures were released, there's a jump in how voters see the state employment situation.

Now, 38 percent say we've gained jobs; 30 percent say we lost jobs; and 29 percent say there's been no change. Gousha pointed out the big jump in how many voters say the state has gained jobs.

"So, the numbers can depend on the messaging of the moment?" he asked.

Franklin agreed.

"The jobs figures released impact responses," he said. "The numbers are certainly driving the commercials being aired."

Voters feel Walker is better at job creation, leading Barrett 48 percent to 41 percent two weeks ago and leading this week 50 percent to 43 percent.

Walker's approval numbers have barely moved since Marquette started their poll in January. The latest results put him at 51 percent approval rating vs. a 45 percent disapproval rating.

"The takeaway is how stable things have been over the year," Franklin pointed out. "We started in January at a 51-46 split, but today there’s hardly any change. Any dips are just sampling noise, but not representative of a big trend or movement."

What it really means, he added, is that voters are unusually engaged and have picked sides.

"There is little evidence minds are changing in significant ways this late in the game," Franklin concluded.

Where there is a lot of movement is in the favorable vs. unfavorable ratings for Barrett. Franklin said we're seeing more of a trend here because Barrett moved from being a local mayor to being a statewide candidate. 

As of this poll, Barrett sits at 41 percent favorable and 46 percent unfavorable. His numbers in January were 34 percent unfavorable and 27 favorable. Still,  Franklin said the differences are relatively small.

Gousha pointed out that turnout is the biggest variable.

"We just don't know because this is unprecedented," he said.

Franklin agreed.

"There was an 800,000 gap between the 2008 presidential election and the 2010 elections," he said. "We are in uncharted territory, but if you look at indicators like in the primaries, we could certainly have high turnout. If you use the sampling model, we could see 2.7 million on June 5."

The Campaigns Respond

Shortly after the MU poll was released, Barrett's campaign issued a statement dismissing the results:

"Over the past five days, a series of polls have been released to the public on the campaign for Wisconsin governor. Most of them show the race either tied or separated by just a couple of points. This includes a poll released today that shows the race in a dead heat, 49%-49%. Only the MU poll is an outlier. The MU poll predicts the electorate to resemble 2010, which is reflected by its undersampling of younger voters, and an oversampling of some of the reddest counties in the state. Yet given the high enthusiasm on both sides, not just on one as was the case in 2010, this expectation is not credible. Quite simply, the results reported today by MU do not sufficiently record surging Democratic enthusiasm. This race will be close, and Tom Barrett will win on June 5th because he will put Wisconsin first, truly focus on jobs, and restore trust and integrity to the governor’s office."

But a quick check of the methodology of the survey reveals that the Milwaukee DMA (the whole of the Milwaukee media market) is comprised of 10 counties, including Milwaukee County, and makes up 32.7 percent of the sample. The City of Milwaukee alone accounts for almost 13 percent of the sample. Other metropolitan areas - DMAs - include Madison (16.8 percent); Green Bay (18.5 percent); and the rest of Wisconsin (19.3 percent).

More, 42.5 percent of those polled identified as Republican; 48 percent accepted the Democratic label; eight percent said they were Independent; and 1.4 percent either refused to answer the question or said they weren't sure.

As for Barrett's campaign saying the survey skews to older respondents, the methodology explains that younger voters typically don't respond to pollsters. Thus, the poll combined census data from 2008 and 2010 to come up with as fair a representation as possible:

In this sample, the population values for age and sex were determined by combining the 2008 and 2010 Current Population Surveys conducted by the U.S. Census in Wisconsin to estimate the
distribution of education for registered voters in the state.

Patch has messages into the Walker campaign and will update the story when we get comment from them.