Would you name your kid Norbert?
Another St. Norbert poll is out. Someone incredibly wise about such things (me) cautioned yesterday when the first results were released showing Scott Walker ahead of Tom Barrett by 9 points, “To my friends on the right looking at the St Norbert’s poll: be happy they supply the cheerleaders for Packers games and ignore the poll.”
I added, “Don’t get me wrong. I bet @ScottKWalker is ahead, too, but I wouldn’t trust a St Norbert’s poll farther than I can throw Mike McCarthy.”
So when the St. Norbert poll says Senator Russ Feingold is only trailing Ron Johnson by two points, you’ll forgive me if I feel justified in being skeptical.
I have no idea what the next real poll is going to say. I’m guessing Johnson is still ahead, and that Norbert is an outlier. We’ll have to wait and see until a real poll is conducted.
The media in this state loves to quote the St. Norbert polls as much as they love Beloit College’s list of things college freshmen never experienced. They love it even more when the poll says something counterintuitive while also telling the media something they want to hear. Feingold has a chance to win? It’s a close race? They love it.
Meanwhile, people “in the know” understand the St Norbert poll is about as accurate as my horoscope. My favorite criticism leveled at the St Norbert poll was earlier this year when Democrat Bill Christofferson noticed something odd.
Here’s one quick measure of how unreliable those numbers are: According to the poll, 93 per cent of the people interviewed plan to vote in the Republican primary for governor, and all but 1% of them plan to vote in the general election.
Remember, these are not people on a voter list they called; just random Wisconsin citizens. And 99% of them plan to vote in November.
Remarkable, since the highest turnout race in memory was the 2008 presidental, when 70% of Wisconsinites voted. In the last race for governor, in 2006, turnout was 51%. In 2002, when Jim Doyle beat Scott McCallum, it was 44%.
Typically, turnout in the primary is about half as much. In 2002, with a hot Democratic primary between Doyle, Barrett and Kathleen Falk, total statewide turnout in both primaries was 22%. Here are the numbers.
So a 93% turnout in September for the Republican primary and a 99% turnout in November certainly would be remarkable — sort of like the sun rising in the west. Or flying pigs.
But these poll results — which claim to have a margin of error of plus or minus 5% — will be reported as legitimate by the news media, who either don’t know any better or don’t care enough/are too lazy to find out the underlying discrepancies.
If you actually want a good perspective on polling, albeit from a Democratic point of view, Christofferson attacked polls by St Norbert and UW-Milwaukee in 2006.