Azimuth Admits Regional Variations in Response Rate May Have Influenced Poll
Dave Nalle, president and founder of Azimuth Polling, who put out poll results a few days back indicating that Ron Paul is leading in Texas, discusses how unexpectedly high response rates in Houston may have aided Ron Paul by Walt Thiessen
(libertarian)
Friday, July 15, 2011
Dave Nalle, whose polling firm, Azimuth, recently released a poll showing Congressman Ron Paul in front of Texas Governor Rick Perry by five points among Texas Republicans, admitted this week, "Maybe we got too many live responses from people who live in certain sections of the state. It wasn't intentional, but it looks from the numbers that we caught a lot more people at home in the Houston area than we expected to, out of proportion with the rest of the state, and that includes the area that Paul represents. We may do another poll to deliberately switch things around a bit to get the Dallas and San Antonio areas better. Of course, San Antonio is very strong Ron Paul territory, so it may not make a difference," he said. "The level of response from Houston was about 27% higher than expected and Austin was about 18% higher than average for the other regions. I think this was a result of timing since the bulk of the calls to those two areas were made on a Sunday afternoon. Henceforth we're going to randomize the calls more for day of the week and time of day. The raw number of votes above the averages for those two regions was only 21, so that's about a 2.3% additional possible margin of variation on the total if we were to count that as an anomaly. Not much, but identifiable."
In an article I wrote about his poll of Republican presidential hopefuls in Texas, I quoted the poll's published description that the voters they targeted were "highly active". I asked him to define what "highly active" voters are.
"In that poll," he replied, "It was a mix of precinct chairs, campaign donors, and multiple-repeat voters in Republican primaries. So at the very least they were reliable Republican voters, but a majority of them, about 55%, were actively involved in party organizations, either in clubs or as precinct chairs. I was able to get lists because I have connections within the Texas Republican Party. I was able to get lists from local Republican clubs and from precinct chairs in those parties. These are people who influence their neighbors about how they're going to vote. So the theory is that we'll get more accurate results from an early poll like this than you'll get from just a poll of the general population. If you poll the general population, half of them don't even know who the candidates are."
I asked him about his company and where the name came from. He said it comes from the astronomical term that describes the position of a star in the sky, as well as its use with artillery, where the azimuth is the adjustment you do to the vertical angle of your artillery piece when you're firing a shot. He thought it was an appropriate name for what his team is trying to do.
Azimuth Polling has its roots from when Nalle was involved in state level campaigns in Texas over the past few years, for whom he did some polling. He has an associate who is an actuary who gets involved in the polling methodology. One of the campaigns was a federal campaign, the other a state senate campaign. He also advised other campaigns informally, and he did work for the Republican Party at that time, including some informational, get-out-the-vote, and limited polling for the party. After the last local election cycle, he noticed that the only company doing any polling in Texas is a company named Public Policy Polling (PPP), based in Raleigh, NC, which he describes as having a very strong left-wing slant. He wasn't pleased with the results they were presenting "because they spin their results pretty heavily". He also claims they've gotten a lot of national coverage within the last six months.
Nalle told me PPP has been paying particular attention to Texas elections, for what reason he didn't know. They've been doing a lot of issues polling, as well as polling on the upcoming senatorial and presidential races. Nalle claims he knew that they only had three guys working for them, although a perusal of their website suggests that they're actually quite a bit larger than that. Nevertheless, his perception is what led him to decide he could start his own company. He knew some college students who could do the work, so given his experience with automated polling systems, he decided to try doing some of that in Texas and perhaps even on a nationwide level.
"If you look at their polls, you'll see a lot of opinion content along with the polling data. It's not really analysis so much as it's interpretation. I didn't want to do that, but I wanted to put out some numbers that were maybe more reliable than what they were putting out, maybe more objective on some of these races. So that's what we decided to do," said Nalle. "One of the things I wanted to do was to go for a slightly different kind of poll than what most people are doing. I've been trying to target people who are more politically involved than just average voters."
Dave Nalle has basically many of the same political beliefs that Ron Paul has. Paul came out ahead in Nalle's poll. So I asked him point blank, "How do we know you didn't fix the numbers?"
He replied, "If I had fixed the numbers, Gary Johnson would have won."
"Oh really! You're a Johnson supporter," I said.
"I like Ron Paul too," he admitted. "But if I could actually choose who would win the election, I would pick Gary Johnson. I'd just as soon have Paul win over anybody else that runs."
"It still raises the question," I followed up, "How can anyone trust your numbers? You don't have a long track record, so how could anyone know?"
"Right," he replied, "Well, as with any poll, they have to make their own decisions as to whether it's reliable or not. What I can say is that I've looked at other polls, and our results are not wildly out of the expected range. In other polls, not just in Texas, but elsewhere, the ratio of votes between Paul and some of the other contestants is similar. In some of those polls, somebody comes in much higher, but Paul comes in second or third in most of them, and the percentage he gets is similar to what we had. The difference appears to be that we did not include anyone who is not running like [Sarah] Palin."
"Why did you leave Palin out," I asked.
"Because as far as I can tell, she is not running," Nalle answered.
"Even though she's been doing all these tours around the country in her bus, and so forth?" I added.
"Right, I think she's just having a good time, promoting herself. I don't see any indication that she really wants to run," he said. "If she gets in the race, then it'll all change, but right now I don't have enough confidence in her to be a candidate."
I said, "Well, you could have asked a second question. If Palin and a couple of others decided they were going to run, who would you vote for? But you didn't ask a question like that."
"We did this pretty fast on a simple basis. We just wanted to get a sense of declared candidates who were running in Texas and get a baseline. If we wanted to make it a full-fledged, across the board, ideal poll, we would have asked different mixes," he explained.
I asked him what Azimuth's future polling plans are.
"We're working on a poll right now which is being done with somewhat more complex methodology for the senatorial race in Texas and another one for the 25th congressional race."
"But how about for the presidential race?"
"We're going to wait a little bit. I'm not sure if it's going to be a Texas poll or if we'll try to do something on a nationwide level. There are a lot of different nationwide polls, and we don't want to do one that doesn't offer something different. We might do a poll on issues, in a nationwide poll, focused not so much on the presidential race, but rather the key issues in the presidential election. One of the things we want to assess, for instance, is how strong social conservatism really is within the Republican Party, with both [Minnesota Congresswoman Michelle] Bachman and [Texas Governor Rick] Perry playing so heavily with social conservatives. That's something I think the media and the major polling networks aren't assessing effectively.
"One of the reasons we poll is to take what I think is going on and find out if I'm right. The [poll's results] will either confirm or make me re-think what I'm thinking. In the case of the Republican primaries, it looks like [candidates] may be playing to the religious right in a way that may damage them in the general election and even in the primaries because they played so heavily to the religious right," he explained.
I took the conversation back to the Texas Republican primary and pointed out that major polling outfits like Zogby, Rasmussen, Quinnipiac, Gallup, etc. will keep doing the same poll over and over again, to find out what is changing if anything. So if they did the latest poll about the Texas primary, wouldn't it make sense to do follow up polls?
"We only have limited resources," he replied. "We're doing these things on spec, pretty much. If someone were paying us to do them, I'd say instantaneously, 'yes'. But since we're just doing this for promotion, we'll probably wait until August to do [a follow up poll]. The C25 and senatorial polls are a little higher priority as far as I'm concerned because those are two races that are very heavily contested. The senatorial race has eight candidates, and none of them is really leading the pack. We're trying to find out what the effect of Lieutenant Governor [David] Dewhurst coming into the race would actually be. And then the C25 situation is bizarre. It's a newly created congressional district with an old number, but the current office holder is not going to run there. He's moving over to the C35 district. It's a rather mixed Republican-leaning but not Republican-dominant district on the outskirts of Austin where there are all of these candidates coming down. Currently there's 10 of them who want to run for it. Most of them don't even live in the district. Under Texas law, you don't have to live in the district you run for. They all think this is the hot ticket to grab, and it's a district that overlaps areas which in the past have elected very libertarian-leaning candidates to office. Half of it is co-terminus with the district that elected [state representative] Jason Isaac, a libertarian Republican."
Nalle is hoping to build his polling business by aiming at statewide candidates to hire their services, and he notes that there has been a massive realignment due to population changes. He mentioned Congressman Ron Paul's district, which is now 1/3 the size that it was, and that the area it used to cover is now going to be covered by parts of five different districts. I asked him if this may have influenced Paul's recent decision not to run for Congress again in 2012 and concentrate instead on the presidential election.
"He said no," Nalle observed, "and my contacts in his office also say, 'probably not'. My guess would be that he would still have won with the newly designed district, although the district number would have been different. But who knows? I understand his desire to focus on the presidential campaign. That makes sense too. What I'm trying to find out right now is, who is going to run for his seat?"
I asked him if he thought there was going to be another line of candidates.
"Well, yeah," he agreed, "for all five of those seats, because they're all changed around with the new demographics. The districts on the coastal area are all different now. The Democrats there are in trouble. Two of the seats they have they'll probably lose. There will be maybe one new Republican and different Republicans in other seats. It's all a big chaotic thing. The other thing is that with the new census, all of our elected officials whose terms lasted longer than 2012, they're all up for re-election, even though their terms aren't over. These are offices like state senators and district office holders whose districts got redistricted, like the railroad commissions, the boards of education, and a bunch of others. They're suddenly all up for re-election. So it's going to be a very significant election in Texas because it's going to shake up the whole government."
I touched again on his earlier comment that he was a Gary Johnson supporter and asked him if he planned to pitch his services to the Johnson campaign, to do polling for them.
"Oh, I'd love to do polling for them," he said with a smile in his voice. "I don't know that they need it. I don't want to say anything out of school about the Johnson campaign, but polling probably isn't their priority. That's not the kind of campaign they're running on the stage right now."
UPDATE: Azimuth followed up a few days after this was first published with an article of their own entitled, "Texas GOP Primary Poll: Secondary Analysis", in which they discussed the possible regional variations in more detail.
Did you like this article? If you did, Thumb It! 6
thumbs so far
The views expressed
in this article are those of Walt Thiessen only and
do not represent the views of Nolan Chart, LLC or its affiliates.
Walt Thiessen is solely responsible for the contents
of this article and is not an employee or otherwise affiliated
with Nolan Chart, LLC in his/her role as a columnist.
