Why Rasmussen polls tilt Republican

Printer-friendly versionSend by email

Those Rasmussen Wisconsin polls just keep on coming, so often that it's hard to believe Rasmussen does it out of the goodness of their heart.

Some Democrats are tempted to tout the latest from Rasmussen, only because it looks better for Dems than Rassmussen's previous polls. It shows Feingold ahead in the Senate race and Barrett tied in the governor's race.

We've been saying for years there's something suspect about the Rasmussen numbers, which always seem to have the Republicans doing better than you'd expect.  Like that famous tower, they lean a little to the right. Couple that with the fact that Rasmussen is a GOP firm and doesn't disclose who's paying for the polls, and you have reason to be skeptical.

Now a far better source and analyst than me has dissected Rasmussen's numbers and found that Rasmussen's sample has 6 per cent more Republicans than it should.

That analysis comes from Nate Silver, the genius who gained fame by predicting the 2008 presidential election outcome almost on the button in 49 of 50 states. On his website, FiveThirty Eight, Silver writes:

The bottom line is this: the sample included in Rasmussen's polling is increasingly out of balance with that observed by almost all other pollsters.

Why does it matter? Because polls, whether reliable or unreliable, accurate or inaccurate, are reported by the news media and devoured by political insiders. Early polls can influence a race by giving a candidate an artificial boost, encouraging supporters and donors to get behind someone who appears to be winning or at least has a shot. Early in a race, polls can, to some degree, become self-fulfilling prophecies.

Here's how it works in Wisconsin: Rasmussen does a poll, duly reported by WisPolitics, the Associated Press, Journal Sentinel, bloggers, and others. The poll, overloaded with Republicans/conservatives, tells us that in our candidate and issue choices we are more conservative than we really are. That feeds the "Dems can't do this because voters don't want it" narrative of the media, and give chicken-hearted Dem officeholders a reason to derail progressive policy and legislation.

That may sound like a conspiracy theory, but it is the way things work.

Rasmussen, under fire for some of its findings which clearly fly in the face of common sense, like extraordinaly high name recognition for unknown candidates, has started including its past track record in Wisconsin elections as evidence of its reliability.

But there is a difference between early polls, when there is no way to verify them with the electorate, and therefore much more room for mischief, and a poll done a week before an election, when there is a real check on election day to show how accurate the poll is. This paragraph by Silver is intriguing:

Citing Rasmussen's success in calling past election outcomes, which is formidable, is also somewhat non-responsive, since their house effect was not so substantial in past election cycles. Moreover, most objective attempts to rate pollsters, including ours, rely on an evaluation of the accuracy of polls in the week or two immediately preceding an election (when pollsters have strong incentives to "behave" themselves). They may reveal little or nothing about the accuracy of polls months ahead of one.

Caveat emptor.