Monday, October 25, 2010

Not Buying Dayton's Lead

A seven-point margin among likely voters seemed a stretch when I saw the Strib's Sunday front page, but then again I long ago dismissed the Minneapolis paper as actually having earned, with all due respect to our aquatic friends, the moniker of "fish wrap." Mitch Berg breaks into the internals of a couple recent polls:
Look at the likely voter percentages in the last few “major” polls”

Strib/”Minnesota” Poll – D+4 (meaning they figure that Democrats will make up four percent more of the electorate than Republicans on election day. 
Rasmussen Poll: D+5 
Now, this is a function of how these polls determine “likely voters”. This formula varies among polling services, but – since it’s a form of science, however imprecise – is hypothetically based on some kind of math, derived from experience.

And what has “experience” been in Minnesota, especially recently?

In the 2008 election, Minnesotans’ spread was D+4.

In other words, Democrats made up 4% more of the electorate than Republicans did.

The pollsters are honestly suggesting that Democrats are going to turn out in the same number as in the Democrat landslide of 2008?

Or that Independents are going to break the same way they did two and four years ago?
Not likely this time around. Of course, as usual, the confidence intervals for Dayton and Emmer are large enough to overlap and provide some CYA when the election turns out tighter than that next week. Curious how that happens every year.

No comments:

Post a Comment


Related Posts Plugin for WordPress, Blogger...

Because Life is Life
and not just on election day