Two Polls; One Interesting, One Bad

By Jim Ellis

Nov. 1, 2017 — Two political polls were released into the public domain yesterday. The first survey is of interest because it already examines a budding open seat Arizona Senate Republican primary, and becomes the first gauge of how former state Sen. Kelli Ward, who had been challenging Sen. Jeff Flake (R), performs in a new situation.

Another in a series of Virginia gubernatorial polls was also published Monday as we begin the last week of the election contest. Here, we question the results’ accuracy.

Arizona Senate

Speculation has been hot and heavy during the past few days, the first time interval since Sen. Flake announced he would not seek a second term largely because of dismal re-election prospects. Culminating the early conjecture are Phoenix-based Data Orbital’s (DO) new conclusions (Oct. 26-28; 500 likely Arizona GOP primary voters) finding former state Sen. Ward leading a pack of potential Senate candidates, but with a margin that suggests the impending Republican primary is anybody’s game.

Looking at the DO results, Ward places first with 26 percent, followed by Rep. Martha McSally (R-Tucson) at 19 percent, and former Rep. Matt Salmon (R-Mesa) coming next at 10 percent, while ex-Rep. John Shadegg (R-Scottsdale), Rep. David Schweikert (R-Fountain Hills/Scottsdale), and Arizona University Regent Jay Heiler score 6, 5, and 1 percent, respectively.

Since Ward has largely been campaigning virtually non-stop for the Senate since 2015 when she began her losing primary challenge to Sen. John McCain, but finished with a respectable 51-40 percent showing, a 26 percent mark against a field without the beleaguered Flake is less than one might expect for her when matched opposite a relatively large group of unofficial potential candidates. This result likely confirms the analysis that her 2016 finish and drawing significant early leads against Sen. Flake is due to Republican dissatisfaction with their incumbents, rather than a groundswell of support rising to elect Ward.

These early results, though by no means indicative of how the early race might break considering the situation’s fluidity, suggest a wide open Republican primary, and one that will take months to complete. With the candidate filing deadline not until May 30 and the nominating elections on Aug. 28, the respective primary campaigns are only in their infancy. Therefore, prospective candidates have plenty of time to assess their standing and can make a clear determination about their ability to win the nomination and general election campaigns.

The Democratic side was not tested in this particular poll. Rep. Kyrsten Sinema (D-Phoenix) remains the only announced Democratic Senate candidate, to date.

Virginia Governor

Quinnipiac University distributed what will likely be their last Virginia political survey before the regular odd-year general election, and all indications point to it being an outlier.

The survey (Oct. 25-29; 916 self-identified likely Virginia voters) finds Democratic Lt. Gov. Ralph Northam leading Republican Ed Gillespie by a whopping 53-36 percent, a 17-point margin that no other pollster has come close to forecasting. This is the fifth poll conducted since Oct. 15, and the only one that has the two candidates separated by more than eight percentage points.

The latter survey, from Hampton University, actually showed Gillespie holding the lead. Therefore, between these two extreme polls, a gap of 25 percentage points exists between their respective final results even though the two surveys were conducted within a week of one another.

Interestingly, the only other poll to show a Northam lead even close to this level was Quinnipiac’s own last data release, a Oct. 12-17 study that found him holding a 14-point advantage, 53-39 percent. The other polls in the current time frame to show Northam ahead beyond the margin of polling error, both arriving at seven-point spreads, came from Christopher Newport University (50-43 percent) and Fox News (49-42 percent).

When looking at Quinnipiac’s segmentation, it is clear that Democrats were over-sampled for their latest poll, in particular. The 37-24-31 percent partisan division among Democrats, Republicans, and Independents, respectively, is out of balance in the Democrats’ favor and clearly does not accurately reflect Virginia voting patterns, particularly in low-turnout, odd-numbered general elections.

Though Democrats have been winning the preponderance of Virginia statewide elections since 2006, their average margin of victory has been just one percentage point over the last eight major statewide elections. The mean was altered when considering Republican Bob McDonnell won the 2009 gubernatorial campaign by 17 points, however. Eliminating the McDonnell result, the Democratic average victory percentage during this same time frame is closer to a three-point spread.

The Quinnipiac skew, notwithstanding, it is still reasonable to believe that Northam has the advantage as the candidates dash for the Nov. 7 finish line, but it is likely nowhere close to the number the Q-Poll projects.