When Pollsters are Shamelessly Biased

By Dave Andrusko

Editor’s note. My family is on vacation. While we are gone I’ll be running articles from the past 12 months that you’ve indicated you particularly enjoyed. Dave

I am as enthralled with polls as the next guy, as anyone who reads National Right to Life News Today knows. But having politically cut my teeth in Minnesota, I know that you need to be acutely aware just who it is you’re reading. Some outlets are far worse than others.

Which brings me to offer a couple of comments on a piece that ran yesterday and is being forwarded by people in or from Minnesota to one another. The headline is “The Big Loser: Media Polling,” written by Mitch Berg.

Berg carefully illustrates how the two dominate polls in Minnesota–the [Minneapolis] Star-Tribune and the University of Minnesota’s “Hubert H. Humphrey Institute”–consistently misrepresent opinion polls on state-wide elections in Minnesota.

The error is especially egregious when the elections turn out to be very close. When these polls vastly inflate the lead of the (almost always) pro-abortion Democrat over the (almost always) pro-life Republican, it can make a huge difference in turnout.

Let’s take just two examples, one gubernatorial, one presidential. In the 2010 contest for governor, pro-abortion Democrat Mark Dayton defeated pro-life Republican Tom Nemmer by one half of one percent. The last HHH Institute poll had Dayton ahead by 12%–41% to Emmer’s 29%.

In 2008, “[T]he day before the election, the Minnesota [Star Tribune] poll said McCain was polling just 37%; he ended up with 44%,” Berg writes. “It overestimated Obama’s support by under a point, calling him at 55% when he got 54.2%. The Minnesota Poll sandbagged Mac by seven points.”

As it happens I’m very familiar with the Star Tribune’s overwhelming bias. Back in 1978 I was writing for a small community newspaper when pro-life Republican Rudy Boschwitz ran against pro-abortion Democrat Wendell Anderson for Senate.

The weekend before the election the newspaper had Anderson ahead by .05%. Actual results? Anderson lost by 16.2%. I wrote a long investigative piece. What I found was really ugly.

The point is simple but important. If, like Nemmer, you know it’s close but the public is told you are getting pulverized, you have that sinking feeling that some unknown percentage of your supporters will be discouraged. That can easily turn a narrow victory into a narrow defeat.

If, like Anderson, the poll tells your supporters it’s nip and tuck, the message is clear: work hard and you could carry the day.

By the way, neither the Star Tribune nor the HHH Institute has ever ‘fessed up to their obvious biases.

One Response to When Pollsters are Shamelessly Biased

  1. Hoffmanla roche small, mississauga, ontario, north america, (pi satrted 4/2000) rechecked 7/2001.

Leave a Reply

Your email address will not be published.