Montana Senate: A Real Race or Simply Manufactured Buzz?

by Stuart Rothenberg July 24, 2014 · 1:59 PM EDT

Maybe you believe in coincidences. I usually do — but not four months from an election.

Almost simultaneously, two different memos appeared from Democratic pollsters insisting the Montana Senate race has closed and the outcome of the contest is very much in doubt.

One memo, by North Carolina-based Public Policy Polling, meets existing standards of transparency and while I have issues with the firm’s conclusions, I was happy with the way the data were presented.

The other memo, by widely respected, Colorado-based Harstad Strategic Research, was dreadful and little more than spin. It fails to meet the minimum standards of disclosure about polls, and devoted more time to promoting the firm’s candidate, appointed Sen. John Walsh, and vilifying Republicans (and the media), than discussing data.

“Montana Senate Race Tightens Considerably” crowed the July release from PPP. Predictably, someone at Daily Kos picked that up and penned an upbeat post about Democratic prospects in the race.

But politics is all about context — just ask the campaign of New Hampshire GOP Senate hopeful and former Massachusetts Sen. Scott P. Brown, which I criticized a few days ago — and the July 17-18 poll of 574 “Montana voters” (PPP’s terminology) isn’t that big of a deal.

The rub of the memo is that, according to PPP President Dean Debnam, the Montana race has closed “over the past eight months,” and is “looking much more competitive now than it did last fall.” Republican Rep. Steve Daines’ lead over Walsh was 17 points in November, but has now shrunk to “only” 7 points, 46 percent to 39 percent.

Of course, last November was a horrible time for the president and his party. The White House was on the defensive over the health care website launch, and that surely contributed to Democratic problems. Using PPP’s November numbers to establish the baseline in the race almost guarantees a closer contest.

Walsh was a sitting lieutenant governor in November, as he wasn’t appointed to Democrat Max Baucus’ vacant Senate seat until February. He was generally unknown, which also depressed his showing in the survey.

You don’t need to believe me. Here is what PPP said in November: “Part of Daines’ lead is name identification. … Because of that, twice as many Democrats and independents are undecided as Republicans.”

While Baucus solidified his seat and didn’t have close races, many Montana Senate races are extremely close. Sen. Jon Tester’s margin in 2012 was 4 points, his margin in 2006 was less than a single point, and Republican Sen. Conrad Burns won his 2000 election by fewer than 3 points.

The 2014 race was guaranteed to “tighten” under normal circumstance, and that normal closing has occurred. If you really thought Daines would win the open seat by 17 points, you had to figure that either Democrats would nominate a third-tier candidate or the GOP wave in 2014 would be huge.

Whatever you think of the PPP release, it shines in comparison with the Harstad Strategic Research memo. According to that memo, written by Harstad’s Andrew Maxfield, “our internal polling has us just five points down,” 43 percent to 38 percent.

The Harstad numbers aren’t all that different from PPP’s, but the Harstad memo is overflowing with messaging, not analysis. In a sense, it is a polling memo without polling.

There are a few numbers in the memo’s second paragraph, but what stands out are two separate mentions of Liz Cheney, one mention of the Koch brothers, one of Karl Rove and two of Eric Cantor. The poll memo also belittles GOP pollsters and the national media.

What does the Harstad poll memo not include? It lacks basic information, including the dates of polls, the nature of the samples (adults, registered voters or likely voters) and sample sizes of the recent to poll which it refers.

I called to get some of the details and was told the most recent survey was done for the campaign, was conducted in mid-July and included likely voters. In other words, plenty of details were still left out, and the information I eventually received was provided only grudgingly after I cited American Association for Public Opinion Research guidelines.

If you are a little confused, you aren’t alone. Why would a legitimate polling firm act like the details of its now partially publicly released polls need to be guarded like the gold in Fort Knox? And why would a veteran polling firm that deals with numbers and rigorous analysis send out a release that was mostly advocacy?

It certainly looks as if PPP and Harstad worked in tandem to try to create a new narrative about the Montana Senate race that would convince political reporters and handicappers — and maybe even Democratic super PACs — that Walsh has a decent chance of overtaking Daines.

The Harstad memo is dated July 17. PPP was in the field on July 17-18, with its memo dated July 21. The core of each memo was the same: The race is tightening and is competitive.

I spoke to a veteran Democratic consultant about this and was told that Democratic consultants know when PPP is going into the field for a specific race. I don’t know whether there was coordination in this case, but the themes of the two memos are remarkably similar.

I have always made a distinction between the Montana Senate race, where the GOP has an advantage, and races in South Dakota and West Virginia, which I regard as lost causes for the Democratic Senatorial Campaign Committee.

But the new PPP numbers don’t lead me to change my opinion of the race one whit. Daines remains the clear favorite. As for the Harstad memo, its lack of basic information, silly attacks on Republicans and emphasis on messaging makes it easy to ignore.

The Rothenberg Political Report rates the Montana Senate race as a Tossup/Tilts Republican contest. [Editor's Note: This was written before the New York Times story about Walsh's plagerism.]