May 3, 2012
Truth in polling
For all sorts of reasons, a bad polling result late in a campaign is the equivalent of a gut-kick in politics. Candidates work like crazy, issue position papers, participate in forums, face editorial boards, appear on talk shows, take all the brick-bats, but all of that work can be offset with one poll even if the poll does not pass the smell test.
Most people think TV spots are biased, but reporters even ask about TV spots if sponsored by a right-wing group or a liberal funder. But a poll, with some unknown director, seems like "science." It is weird, but when statements are made that the "margin of error is 3.7 or 4.6" many people conclude that the poll must be accurate if you just subtract the margin from the results reported. The margin-of-error magic. Not unlike the weather reporter giving you the probability of rain thanks to Doppler radar. It makes it appear to be scientific and unbiased. Think again.
Milwaukee Journal Sentinel reporters, who are sticklers for truth in all things irrelevant, with their child-like "pants on fire" award for a whopper told by a candidate or a staff person, ask few, if any, serious questions about a poll. If they did ask they didn't bother to tell us the results.
About a year ago, we discovered that the far-right Bradley Foundation front, the Wisconsin Policy Research Institute (WPRI), made a deal to "partner" with the political science department at UW Madison. The Poli Sci department agreed to conduct WPRI polls with questions provided by WPRI. WPRI and Political Science agreed that the agreement and all polling would be outside the reach of open records laws. Incredible. Reporters who routinely use Open Records to get information wouldn't think of asking Marquette Law School about their brand new polling operation. Keeping all the relevant data outside the reach of the public is not pants on fire it is "building burning down!" And they agreed that a Milwaukee Journal Sentinel reporter would have first crack at revealing the poll's results. A scoop so to speak. We got this information under an Open Records request. We asked questions like Who is paying for the poll?
The poll released yesterday ignores the earlier unholy alliance between the Journal Sentinel and the Bradley Foundation's WPRI. Why raise that now since the partnership crashed and burned when they could not operate in scret? Well, guess who was with UW Political Science when we exposed their agreement with WPRI--Yes sir, Charles Franklin. Who directed the so-called Marquette University Law School poll? Yup. Charles Franklin.
There has always been a question about cell phone calls vs. land lines. (Young people use cell, old guys still like land lines.) Franklin simply asserts that the poll included cell phones! Did the pants on fire reporters ask how Franklin got cell phone numbers? How many? I get plenty of cold-calls on my land line never on cell phone.
Franklin says the resuts are even better for Barrett now than in the March poll conducted by Franklin. He doesn't mention that Barrett was not even a candidate for governor when the March results were publlished. Who knows what that means?
In addition to all that, keep in mind that Marquette is outside the reach of Wisconsin's open records law! So, Franklin, described as "visiting" professor, or "on leave professor," or "director of the Marquette Law school poll." A person close to Wisconsin politics described Franklin as a gift from Alabama to Wisconsin. Who is Charles Franklin? Is that too tough for reporters to ask?
The poll tells us all sorts of nonsense such as "Have you stopped talking politics in your family?" How about this: Who is paying for the poll? Is Franklin taking the money out of his pocket? Is Franklin on someone's payroll? Does he have a contract with Marquette? These questions are important because a Frankin poll could determine the winner.
post a letter about this blog »
People believe the polls they want to believe. They take delight in any result that bolsters their skewed belief system. Let polls be what they are, entertainment.
-Pietr Haikuu | Hurley, Wisc. | May 3, 2012
These are important questions which most of us don't think to ask. Thanks for staying on this issue.
-janeofdane | Madison, WI | May 3, 2012
Last night Rachel Maddow cited the Marquette poll to make her case that Walker's spending is not influencing the election. Is she another who finds the data to fit the story?
-Maria Caliente | Middleton, Wisconsin | May 3, 2012
Thanks for making this very good point, Ed. Anything related to the Bradley Foundation, including professor Charles Franklin.
-Andy Olsen | Madison, WI | May 3, 2012