December 10, 2010

Newspaper poll for amusement purposes only

Eight hundred and ten eight million dollars worth of investment in public infrastructure fled from the State of Wisconsin yesterday, observes the Milwaukee Journal-Sentinel, "thanks to the opposition of Governor-elect Scott Walker and an apparent majority of State residents."

Yet situated directly left of this claim is a reader poll, showing 64% disagreement "with how Scott Walker has handled the high-speed rail issue." At the moment there's around 16,700 respondents, or 22.3 times the size of any McClatchy-Marist or Rasmussen sampling.

The latter are reported dutifully in the weeks preceding elections and tend to have the effect of influencing those outcomes, as many voters are inclined to support candidates whom they believe will win.

The Journal-Sentinel's refusal to acknowledge the apparent massive unpopularity of the governor-elect's self-congratulatory "victory" (the paper endorsed career politician Scott Walker for the top executive office) undermines the empirical reliability of a previous survey, which showed "shoveling the driveway" edging out "skiing or snowboarding" as the favorite snow activity of Wisconsinites.

Some political scientists believe that only the most ideologically committed activists respond to reader surveys, and this makes them even more skewed than are Mr. Rasmussen's toward Republicans.

Meanwhile the Journal-Sentinel's allegedly economics-minded right-wing calumnist Patrick McIlheran rejoiced in the loss of several thousand construction jobs with an emphatic "Good riddance."

The victorious governor-elect is expected to introduce "tort reform" provisions he suspects will compensate Wisconsin for the shortfall.


gnarlytrombone said...

22.3 times the size of any McClatchy-Marist or Rasmussen sampling

The sample size in this case is absolutely meaningless in determining accuracy because there's no way to know the distribution. And the potential biases are legion - Internet users, JS readers, ability to surf the web at lunchtime - not just the activist skew. That's what makes polling so expensive to get right (and in turn why Rasmussen cuts so many corners). These polls really are just for amusement.

I'm also of the mind that polling is a very poor way to ru(i)n a railroad, even if it's statistically sound. This would be a great alternative.

illusory tenant said...

I wonder if the J-S's own "apparent majority" is derived from WPRI's presentation of the connection as a fancy bus route.

Pete Gruett said...

The JS's editorial is an exercise in cowardice. They talk about the democrats failing to "make the case" for rail as a reason for the public not supporting the project. This is a polite way of pointing out that the public swallowed Walker's BS with plenty of help from the JS.

When the original survey was taken, Walker was claiming he'd spend the money on roads. How many people would've opposed the project if they'd known the alternative was sending almost a billion dollars out of state? Also, how many of them knew that Madison and Dane County
had offered to pick up some of the operating costs McIlheran was so afraid would eclipse the initial grant in 1000 years or so. Apparently, Walker was too busy to return calls from the mayor of the second largest city in the state.

illusory tenant said...

How many people would've opposed the project if they'd known the alternative was sending almost a billion dollars out of state?

Good question.