CAAAD 2012 Travis County Democratic Primary Poll

DA Race (2.00 / 1)

I'm wondering if there's been any other public polling on the DA ace. I had hoped that Charlie Baird was much closer to taking over Rosemary Lehmberg's lead.

---

by: persiancowboy @ Mon Feb 06, 2012 at 17:01:45 PM UTC

Re: DA Race (0.00 / 0)

There has been other polling in the DA race, but not publicly. I asked around and as best I can tell you it was done towards the end of 2011 and as I'm told, it generally was similar to this one.  FWIW.

Please read the Community Guidelines and How to Rate Comments.

---

by: Karl-Thomas Musselman @ Mon Feb 06, 2012 at 21:42:56 PM UTC

[ Parent ]

There Is other polling (0.00 / 0)

There is other polling, including Baird campaign polls which have previously been released by the campaign. No poll we have seen, either conducted by the Baird campaign or a third party, reflects the numbers in the poll above.

On November 15, the Baird campaign sent an e-mail memorandum to its donors outlining the results of a poll taken November 8-10. We will excerpt some of what we sent our donors below:

If the race were held today, Charlie Baird stands at 20.3% and Lehmberg stands at 23.6% with the remainder undecided--well within the margin of error.  

We are besting or essentially even with Lehmberg in nearly every demographic across Travis County in a race that is neck-and-neck and could shape up to be one of the closest countywide Democratic Primary contests in Travis County in the modern era.  

Since our first poll about five and a half months ago, Lehmberg has fallen almost seven full percentage points, while we have increased by nearly ten.  

When examined more closely, the results are even more promising for Baird:

•Among women, between our first poll and our most recent poll, we overcame a  29.6 point disadvantage to pull to within two points of Lehmberg in a  demographic group that should theoretically be predisposed to vote for a  female candidate.  

•Among African American voters, Lehmberg lost a tenth of a point and we pulled  up seven points and are virtually tied with her among these voters.

•Among Latino voters, Lehmberg plummeted 16 points while we increased our  numbers by nearly 17 points.

It is abundantly clear to us that, when voters are familiarized with Baird's message, Lehmberg's message and, when given a choice between the two, voters choose Baird overwhelmingly.

In addition to the internal campaign polls, we are aware of at least two third-party polls, both of which reflect a large number of undecideds with Baird and Lehmberg both within the MOE. In one of these polls, Baird leads, in another, Lehmberg leads, but both are within the MOE.

There is a third poll we are aware of conducted by a third party where Lehmberg bests Baird by four points, but it was also a robo poll.

While CAAAD's polls have been more accurate in prior years, voters should be aware that CAAAD's previous primary polls have been conducted by at least two other firms, not Southwest Opinion Research. That data is available here on BOR by clicking the hyperlinked "multi-year tradition" text above and reviewing prior polls.

For example, the 2010 CAAAD Primary Poll conducted by KCZ Consulting featured a separate MOE (Margin of Error) for each question. This is a standard practice in robo polling because the drop-off is typically quite significant the longer the poll is. With the 2012 poll, the MOE is uniform, which raises several questions, because it is highly unlikely that every person polled responded to every question. And, if they did not, given that the DA's race question was first, we would question whether the pollster used some type of elimination process in order to come up with an end result with a uniform MOE across all questions. Typically, the MOE for the first question would be the lowest and the MOE for the last question would be the highest, as the number of respondents would decrease with each question. Here, we are left to assume that, somehow, every single respondent answered every single question.

To get a little better understanding of why robo polling in general is not always the best gauge of an end result, we would encourage folks to look at CAAAD's 2010 Primary Poll and the end election result for 299th District Court Judge. In that race, Leonard Martinez more than doubled his numbers from the CAAAD Poll to Election Day. While Mindy Montford remained essentially within the poll's MOE, Sage jumped significantly between the poll and election day (which was not a particularly large window of time). Robo polls typically undersample minorities and young people for various reasons including use of cellular phones as the "main" household phone, and the fact that there are often less accurate phone numbers available in commercial voter data for minority and youth households. In Austin, in particular, the under 35 population is a very mobile population because of UT.

At first blush, an undersampling of Hispanic voters would appear to be part of the reason for the significant difference between the end result of the 2010 primary and the CAAAD 2010 Primary Poll.  

---

by: Charlie Baird Campaign @ Tue Feb 07, 2012 at 19:00:32 PM UTC

[ Parent ]