Editorial: Never take polls as gospel

Talk about it

A Minnesota Public Radio-Humphrey Institute poll released last week was touted as excellent news by the campaign of Democratic gubernatorial candidate Mark Dayton.

The Dayton camp had an obvious reason for its enthusiasm. A month ago, the same poll showed Dayton and Republican candidate Tom Emmer each getting 34 percent of the vote, with Independence Party candidate Tom Horner at 13 percent. Last week’s poll results, however, have Dayton leading Emmer 38 to 27 percent, with Horner moving up slightly to 16 percent.

What does this mean exactly? Well, a lot .., and ultimately nothing, too.

Examination of the poll reveals that Democrats have significantly increased their level of interest in the Nov. 2 election. Perhaps that’s true, but one can also draw the somewhat logical conclusion that as an election draws nearer, interest in any given party’s candidates should grow.

The poll also shows that one in five Republicans — about 22 percent — now plan to vote for Horner instead of Emmer. That could well bode trouble for the GOP candidate, but there’s also another big number out there — an undecided percentage of nearly 20 percent.

Here’s another number that shouldn’t go unnoticed: 750. That’s the number of likely Minnesota voters surveyed for the poll from Sept. 22-26. Minnesota Public Radio stated the poll “has a conventional margin of sampling error of plus or minus 3.6 percentage points,” but 750 people — though it may be a perfectly legitimate sampling by poll standards — by no means encompass an entire state.

We still have an entire month of campaigning left, and October can be notorious for surprises. (Note Mike Hatch’s final-week comments to a Forum Communications Co. reporter in 2006 that by no means helped his chances.) And remember, people still have to cast those ballots on (or by) Election Day.

To quote Yogi Berra, it ain’t over ‘til it’s over.

Tags: opinion, editorials, politics, campaign, polls