Friday, September 7, 2018

Topline Thoughts on Dueling County Executive Surveys


I am, of course, talking about the surveys featured in Josh Kurtz’s recent Maryland Matters article, which can be found here.

-       Call me old-fashioned but if political surveys aren’t conducted using candlestick phones with exchanges like Klondike5, they are garbage.
-       Seriously, robocall surveys are rarely a good idea.  Don’t be cheap, hire trained live interviewers - especially if you want to gather insights beyond surface favorability and horserace numbers.
-       I would like more information on how both groups defined/screened-in “likely voters.”  I had the best screening question for that.  The key is to make it socially acceptable for less-than-likely voters to be completely honest with you.  Of course, it works best when using live interviewers…
-       Let’s start by looking at the survey fielded by the Democratic-affiliated independent expenditure group, then we will turn to the Republican pollster’s findings:
o   Ball down by only 2.8%?  Frankly, that is a better-than-expected showing in my estimation.  If so, it would be due to an effective ground game and a generally favorable political environment.  Ball hasn’t launched his air game yet while Kittleman has been on TV, at least with cable spots, for months.  The good news for the challenger is that Kittleman is nearly maxed out in terms of Name ID while Ball, as is noted, has more “room to grow.”
o   Kittleman’s favorability margin (2.6:1) is not that much better than Ball’s (2.05:1). Ball’s overall favorability percentage should click up as more voters become aware of his record and vision for the County.  That said, as the “comparative” communications kick in (not necessarily from the candidates or their campaign teams) I would imagine that both Kittleman and Ball will see increases in their unfavorable numbers.
o   I am assuming Ball’s “unaware” numbers are higher (by a statistically significant margin) in the 1st and 5th County Council Districts.  The true battleground is going to be, again, the First District, but it wouldn’t be a waste of time for Ball to visit the West to chip at the margins.  There is a strong Democratic team out there running in the 9th and the 5th CC…some joint appearances and targeted door-knocking could help all concerned.    
o   Beyond that, it might be worthwhile for the Ball campaign to spend time and resources in those precincts in District Three where he may be less familiar (compared to Columbia where his Name ID should be strongest) …areas like North Laurel, Savage, and Jessup.  I haven’t seen the survey data, but if there were precincts with a sizable number of Democrats who would be inclined to vote for Ball, but they don’t know a ton about him at the moment, it would be in the First and in the aforementioned swathes of the Third.
o   Meanwhile, beyond trying to run up his numbers in the 5th, Kittleman is clearly going for a W in the First…and his advertising demonstrates a willingness to try to peel away votes from Democratic strongholds (most notably the Third and Fourth Districts).  Ball can’t forget about the base since Kittleman’s campaign is using Pioneers and traditionally Democratic surrogates (such as educators) in his communications.
o   Back to the survey: 51.6% Right Direction/20% Wrong Track.  Sounds about right.  This means that voters – at present – do not see a need for radical change.  They can be convinced that one candidate offers a better approach than the other…but a theme akin to “it’s midnight in Howard County” would fall flat as it does not align with the electorate’s perceptions.  There is no stomach for revolution, but smart reform can be a winning message.
o   It is clear the “most important problem” question was presented using a closed-ended framing.  In that set of circumstances, I buy “cutting property taxes” emerging as the top priority but I wonder about the intensity of feeling…and how the results would have been different if the respondent were asked the question in an open-ended manner.  If that had been done, something tells me that controlling development would have been neck-and-neck with taxes.  In any event, I would love to see the Unaffiliated & Lean D breakouts on that question.
o   Turning to the GOP survey, Kittleman’s favorability numbers are around the same as they are in the Chism survey (64% compared to 60%); the big difference is in the unfavorable numbers (23% in the Chism poll, only 11% in the Burton-conducted survey).  This has me believing that they employed different 2018 General Election turnout models…and perhaps Burton was using a model similar to 2014, which was a bit of a high-water mark for the GOP.
o   Comparing the Ball favorability numbers…again it was 32% fav – 15.6% unfav in the I-E study fielded by Chism whereas it was 29% fav – 7% unfav in the Burton study. The favorability margin for Ball is better in the poll released by the GOP-aligned firm (2.05:1 for the former compared to 4.1:1 for the latter).  Burton’s sample is generally less aware of Ball (53% Name ID compared to 73% Name ID on the Chism survey).
o   The Right Direction/Wrong Track numbers also reveal a more optimistic electorate, 71% Right Direction/19% Wrong Track for the Burton survey compared to 51.6% Right Direction/20% Wrong Track on the Chism poll…both are very similar on the Wrong Track numbers but Burton’s Right Direction numbers seem…rather high to this author/pollster.
o   Again, Ball’s share of the vote on the head-to-head ballot test is similar on both surveys (35.2% on the Chism survey, 37% on the Burton poll).  The big difference is with Kittleman’s numbers…38% from the study conducted by the D-associated I-E organization and 52% from the survey fielded by the Republican polling firm.
o   If you believe the Burton numbers, the undecideds currently comprise only 11% of the HoCo electorate.  I would think the number would be closer to 25% at this stage of the contest, understanding it will fluctuate…then decline as Election Day draws nigh.  The Chism study has 27.8% Undecided at present…that feels closer to where reality is…assuming we are measuring apples to apples and that both ballot tests were “initial/cold” and not “informed” which occur on surveys after information about one or more candidates is shared…
-       Bottom line: do I think Ball is down 15%? No.  Down only 2.8%? That seems closer than I would expect at the moment. Once the air game is joined and Ball’s messaging reaches more households that are less likely to eat and breathe politics, I think he closes the gap.  Prediction: Kittleman and Ball will be in a statistical dead-heat by the first week of October.
-       While I trust robocall surveys less than live interviewer polls, I think this is an example where this robocall poll might have been more accurate in capturing the mood of the county and the state of the CE race.  Without having seen the internals, it seems as though their voter turnout models are different.  The Burton poll, based on the released numbers, appears to be based on a GOP best case scenario model whereas the Chism survey seems to assume heavier turnout from Democratic-aligned constituencies…which I believe to be a reasonable assumption for 2018.

In solidarity.

No comments:

Post a Comment