Unlikely Voter

Tweet

By request (requests always taken via the Contact page, but in general know that individual House race polls are few and far between), I’m taking a fresh look at the Maine Governor’s race.

This is a three way race between Republican Paul LePage, Democrat Libby Mitchell, and Independent (and former Carter and Muskie staffer) Eliot Cutler. The polling has increasingly favored LePage thanks to the split race, but what’s going on with the new Critical Insights poll, is what I think everyone wants to know.

Rasmussen Reports showed LePage +18, and Public Policy Polling had LePage +14, so the convention wisdom has grown that the split race is handing the seat to the Republican. But enter the Critical Insights poll, which shows Mitchell 30, LePage 29, Cutler 9 (MoE 4.9). That’s nearly tied, and I show it a 54/46 split in favor of the Democrat, instead of every other poll’s blowout for the Republican.

What’s going on? I think the key is what we see in Slide 3 of the in depth Maine Today presentation: “Results presented in this report are based on statistically weighted data to reflect the gender, age, and geographic distribution of Maine Voters.” So without knowing the unweighted results, we don’t know if the random sample was any good at all.

I’ve said it before, and I’ll say it again: If you can’t take a proper random sample, then you’re missing the entire point and justification of scientific polling, and ought to take up knitting instead. Because you’re just out of your league if you think that massaging the numbers is a substitute for a random sample.