21 October 2016
Why so few public election polls in Montana in 2016?
Between Labor Day and 21 October 2012, five polls examining the gubernatorial election in Montana were publicly available. Another four were released before election day. Flathead Memo published a graph of this poll on 5 November 2012, the day before the election, and a graph of the polls in the election for U.S. Senator.
Four years later, only one publicly available poll of the contest for governor is available, the Mason-Dixon poll commissioned by the Lee Newspapers, which are still still reporting the results. Has only one poll been conducted? Or have numerous polls been conducted, but their results kept secret from the public?
Montana Cowgirl provides the answer:
There has been a ton of polling by the various groups, parties and campaigns in all of these races and much of it comes to the attention of the Cowgirl Blog. Dozens of polls since the spring tell the following story: Bullock had a solid lead throughout the spring and summer, as much as 8 points on average. It then narrowed when Gianforte began spending big on TV, with a single poll even showing Gianforte up a point but the mean average of polls still showing Bullock up about 3 or 4. Then in the last week the spread widened considerably, with Bullock now averaging about 6 point lead. Factoring a roughly three point bias, the Mason Dixon poll is what you would expect it to be. But again, it is a Republican leaning poll, and always has been.
In 2012, the first Mason-Dixon poll reported Bullock was leading Hill by one point, 44—43. The second reported Bullock was trailing Hill by three points, 46–49. All polls have a house effect, or bias. After each election, reputable pollsters review their protocols and make changes to adjust for biases. Mason-Dixon’s current bias, if any, may be three percent Republican, but because pollsters tweak sampling and weighting after each election cycle, the biases of the past are not a sound basis for predicting the biases of the future.
The bigger question, however, is why that ton of polls have been locked up this year, their results denied to the voting public. My hunch is that on the progressive side, the polls are reporting a steady rightward drift, a drift that predicts defeat for the Democratic candidates for Montana Secretary of State, Attorney General, Auditor, and Superintendent of Public Instruction. If so, is that information being suppressed because the campaigns fear it will demoralize campaign staff, volunteers, voters, and donors? I’m not ready to draw that conclusion, but I cannot ignore the possibility that such Machiavellian machinations may have occurred.
I hope the Lee consortium commissions another poll for late October, and releases the results, and the questionnaire and crosstabs, on a single day.
As for that ton of polls in the lockbox, free it. Let the people whose vote you seek know what you’ve learned about them.