So the Rasmussen poll has made the rounds (so to speak) in the last few days showing Mike Rounds leading the Senate pack with 44%, Rick Weiland at 29% and Larry Pressler at 18%. 7% are undecided, and 2% are someone else. This poll shows that Rounds is firmly planted below 50%, and that the two ostensibly centrist/liberal candidates put together exceed Rounds. While Rounds can only lose voters, sixth months of campaigning and a larger campaign budget than Pressler offer hope for growth to Weiland, who, with a little help from the DSCC, could hit the airwaves and start grabbing those centrist votes, as he tours the state and spreads his populist message. Rounds’ campaign is exclusively televisual – so why not hit Rounds’ turf (TV) and even the odds?
Of course, the real story here (besides Rounds’ inability to even round up to 50%) is the fact that Rasmussen (prophets of Presidents Romney and McCain) did not include Gordon Howie. Don’t assume they didn’t know that Howie was an option for the poll – Howie was included in this general election matchup poll a month ago from Survey USA. Do our state’s great political reporters detect that nuance, and remember that not including someone’s name on a poll means one cannot draw conclusions about that candidate and the race? Hell, no! Here’s David Montgomery’s slapdash reasoning:
The poll found Rounds getting 44 percent of likely voters, to Rick Weiland’s 29 percent and 18 percent for Larry Pressler. Seven percent are undecided and 2 percent pick “some other candidate” — the only option there being Gordon Howie.
It just so happens that correlates closely with SurveyUSA’s poll for the Aberdeen American News, KSFY-TV and KOTA-TV. It finds Rounds at 44 percent, Weiland at 30 percent, Pressler at 17 percent and Howie at 3 percent among likely voters.
Two important points:
1. Survey USA’s poll was conducted from 5/6 to 5/10. Rasmussen’s was conducted from 6/4 – 6/5. This means that Rasmussen, a right-leaning firm with historically unreliable stats conducted their poll after an extra three weeks of Stace Nelson’s relentless campaigning against Mike Rounds. To not include in the most recent poll Howie, Nelson’s ideological successor (and endorser), after the three final (and worst) weeks of Rounds’ primary race, renders the poll meaningless.
2. “Some other candidate” does not mean Gordon Howie. It means “some other candidate”. Perhaps the people who took the poll might have selected Howie, or would select a candidate described as far-right independent. Without that candidate being offered to them, some conservatives probably chose Rounds as someone they would probably vote for.
Montgomery’s analysis is as fundamentally flawed and simplistic as the poll. The fact that he surmises that one set of numbers equals another set from three weeks prior is lazy political reporting and reveals that he is unable to understand the nature of numbers, much less the nature of polls.
Let’s make something clear: When people are put on ballots and polls, that makes them eminently more likely to be selected by voters and poll-takers than if their name isn’t there.
And once again, to be even clearer: You cannot use data from a poll from three weeks ago to make assumptions about a poll from this week.