TRAHANT REPORTS—A new poll in Montana shows Denise Juneau trailing Rep. Ryan Zinke by a wide margin. “Despite record fundraising and polling solid support among Democrats, Juneau, the current state Superintendent of Public Instruction, will need to make big moves in the final weeks of campaigning to close the gap with the freshman Republican, analysts said,” in The Missoulian. Lee Newspapers commissioned the survey. “Among registered and likely Montana voters polled, 40 percent said they’d cast their ballot for Juneau and 53 percent for Zinke. Only 6 percent reported being undecided in the race, and the poll has a margin of error of 3.2 percentage points.”
I’m skeptical and here’s why: The best way to read polls is to look at a number of the surveys. The best data comes from several rounds of snapshots because you can throw out the high, ignore the low, and get a sense of the electorate. This is a single poll that seems to me to fit the outlier category. The 15-point gap is just too wide. Juneau’s own polling shows her trailing by 3 points. That’s probably an outlier, too. A Zinke poll shows an 11-point lead. (Previous: Juneau’s Debate Answer: First Get Elected.)
Who’s right? We might have to wait for Election Day to know that answer. But the good thing about any polls is that it tells candidates where they ought to spend more resources. And, according to Lee’s survey, that’s Independent voters. “The numbers on Independents parallel those in the presidential race, where Trump leads Hillary Clinton 50-28 among those voters. Another 13 percent said they would vote for Libertarian Gary Johnson.”
There are two other reasons why I am skeptical about this poll. The first is that the results so closely match the final result in 2014 when Zinke defeated John Lewis for the open House seat by a 55 to 40 percent margin. But that was an off-cycle election and only 373,831 people voted (or 55.44 percent of those registered to vote). This is a presidential election cycle. In 2012 the Montana turnout was 72.4 percent. According to the Lee newspapers, “The Mason-Dixon statewide poll included calls to landlines and cell phones October 10 through 12. The results included 1,003 Montanans who reported they were both registered and likely to vote. The number of people polled in each county is proportionate to the historical voter turnout. The margin of error for statewide figures is plus or minus 3.2 percentage points but higher for numbers broken down by gender, region, or party.”
Ah. I would like to know more about that last sentence. Of course American Indian voters won’t be counted or statistically relevant. We live in the land of The Other. Damn it. And a poll of a thousand people cannot accurately reflect what Native Americans are thinking or if we will vote. But what if the statewide turnout remains near 72 percent and the turnout from Indian country is 90 percent? That still would not be enough to overcome a 15-point gap.
But what about the numbers broken down by gender? Lee’s report implies that Juneau has a slight lead (the phrase used in the story: “48 percent of female voters said they support the Republican.”) There are two things here: First this survey doesn’t reflect the rising discontent with the Republican presidential ticket; and, more important, it doesn’t measure the intensity of women who will vote. In other words: What if women vote in higher percentages than men? Nationally that’s already the case. Women out vote men by nearly 4 percent. And that number is growing. What if in Montana that gap shoots up to 8, 9, or even more?
As the blog FiveThirtyEight reports: “Men are treating 2016 as a ‘normal’ election; women aren’t.” The post reports a huge gender gap in the presidential race. “We haven’t seen anything like Clinton’s 20-point lead over Trump among women in decades. The last time women favored either party’s nominee by more than 20 percentage points was in 1972, when Republican Richard Nixon crushed Democrat George McGovern among both sexes. The only Democrat ever to win women by more than 20 points was Lyndon Johnson in 1964 — also in a blowout. Four years ago, President Obama carried women by only about 12 points. Even when he first won the White House, in 2008, by about double his 2012 margin, his margin among women was only 14 points.”
I also don’t think you can discount how a blowout election could impact states like Montana. Johnson in 1964 is a good example because he carried all of the Western states (except Arizona). A one-sided election often discourages the losing side, making it more difficult for other candidates from that same party.
My guess — and this certainly runs counter to the poll I’ve been citing — is that this election will be close. It’s Juneau’s history. She only won her last election by 2,231 votes (out of 468, 563). It also fits Montana’s history during a presidential election year. And one additional factor that’s impossible to poll: The number of people who’ve already voted (something that is especially critical in Indian country).
This election is unlike any other in our generation. So the only poll that matters will be that November surprise.
Mark Trahant is the Charles R. Johnson Endowed Professor of Journalism at the University of North Dakota. He is an independent journalist and a member of The Shoshone-Bannock Tribes. On Twitter @TrahantReports.