October 2, 2006
It was early evening, Monday Sept. 18. It was the day after we finished off an issue of NPT, so I was taking it a bit easy. My boy Dante, 17 months old, was napping, and I was lying on the couch with him, trying to decide whether to sleep some more or wake all the way up and watch Monday Night Football.
Just as I decided to watch the game, the phone rang; I grumbled and picked it up. The caller opened with a greeting, and said she was doing surveys about the upcoming elections. Drowsy and annoyed, I almost hung up, thinking at first it was a telemarketer. As it turns out, she sort of was.
The issue of push polling came up when I did a story in June on the public surfacing of the state Senate race between incumbent Eliot Shapleigh and Dee Margo. Ouisa Davis, who writes a regular column for the El Paso Times, and said she has been a supporter of Shapleigh’s, received a telephone call that she interpreted as a push poll. [story]
A push poll, simply stated, is a series of slanted or false statements delivered in the guise of a poll, and is designed to plant unfavorable information about a candidate. Actually, there is a much more specific definition of a push poll, which I’ll get to shortly. Suffice it to say they have been thoroughly discredited and are considered highly unethical. And something told me this might be one.
So, shaking off my fog, rolling Dante off my belly and onto the couch, I sat up and said I’d take the poll, and I started taking notes.
The poll started out with some general questions: When did I vote last, did I vote Republican or Democratic, etc. She asked me to rank the most important issues facing the Texas Legislature.
She moved on to asking my impressions about people, ranging from Laura Bush to the gubernatorial candidates to El Paso County Sheriff Leo Samaniego to, you guessed it, Eliot Shapleigh. The girl asking questions pronounced it Shapely. I thought that was kind of funny. She sounded very young.
Then she got into the meat of it, reading the state Senate candidates and their affiliations. Do you think Eliot Shapely deserves re-election, or do you think you should give someone else a chance? she asked me.
While answering the questions, I tried to put myself into the mind-set of a conservative Republican, and answer how they would answer. But that scrambled my brain, and I started slipping into other roles, so sometimes I answered in the role of, say, a Libertarian, and sometimes as myself, so I’m sure when they tab my results they’re going to be scratching their heads wondering what the hell kind of political outlook I have. It was kind of fun to imagine how I might answer if I was in this or that political box.
It got a bit more interesting as she read a series of statements about each candidate, and asked me to respond whether the statement made me more or less likely to vote for that person.
So I was told Eliot Shapely “is a lawyer who served in the Texas Senate for 10 years,” and that Dee Margo “believes it is important to be involved in the community.”
“As a Republican with bipartisan support” Dee Margo would be able to work with state leadership, while Eliot Shapely “has not been effective” in getting funding for key projects such as the medical school.
Dee Margo’s "successful career in the insurance industry" makes him more likely to be able to deal with health care issues. Eliot Shapely is so difficult to get along with even his Democratic colleagues won’t endorse him.
There’s more, but you get the idea. If you’re interested, read the whole thing here. [survey questions]
The girl was very nice, and she sounded cute, and toward the end of the deal, I even flirted with her a bit as she asked for my demographic information (that bit about demographics is important, as you’ll see in a minute). I asked her if she was calling from El Paso. She said Houston. I said too bad you’re not in El Paso, and she giggled. The whole thing took about 20 minutes.
When I hung up, I was sure I had just experienced a push poll, especially as given the definition published in a previous Newspaper Tree column. [link]
“Voters can identify push polls by listening closely to the content of the questions and how they are asked. Legitimate political surveys attempt to assess the strengths or weaknesses of a candidate or policy by asking a wide range of questions worded in a clear and neutral fashion. Legitimate polls avoid wording that will bias subjects toward or away from a particular point of view. Push polls, on the other hand, often begin with a series of neutral questions and then about halfway through the survey begin to ask loaded questions, full of negative information about a particular candidate or issue,” stated the February 2005 column by then-District 2 candidate and now city Rep. Susie Byrd.
So I contacted the American Association of Public Opinion Research, a leading organization that has condemned push polling.
Dr. Tom Guterbock, AAPOR Standards chair and Director of the Center for Survey Research at the University of Virginia, said the public understanding of a “push poll” is not always in line with the strict definition used by his organization.
“In the popular mind a push poll is anything that is a persuasive poll,” Guterbock said. “The problem is there are legitimate research objectives where you do try to change opinion. ... The problem is that when these are done, we get many complaints about these kinds of polls. They make people uncomfortable and it detracts from the appearance of objectivity from the survey researchers.”
The key elements that define a push poll are duration, the number and type of questions, not collecting demographic information, and the number of calls made. A typical poll might sample hundreds of people, while a push poll would contact thousands of people. A push poll also is short, with a handful of comments disguised as questions.
“Contrast that with a ‘message-testing’ poll, which is a legitimate form of research, although it has its problems,” Guterbock said. Consultants use message testing to figure out which ads will work best.
“Frankly, there are things that suggest to me it was a real message-testing poll being used for research, I guess by the Margo side. And there’s nothing that says that one party’s polls need to be objective if they’re for internal purposes. But no one should be doing polls of the public that leave a lot of voters feeling upset or deceived."
I was asked for demographic information. And a source with knowledge of the survey put the number of people called at 300.
Elizabeth Margo, Dee Margo’s campaign manager, said “Dee is running a positive campaign. The Dee Margo team is not in favor of negative campaign tactics and we look forward to continuing this race with a positive message for El Paso.”
Shapely, err, Shapleigh, declined to speak with me for this article, citing concerns over the publication in NPT of commentary on the race by David Karlsruher, whose fiancée was Dee Margo’s previous campaign manager before being replaced by Margo’s daughter-in-law, Elizabeth Margo.
Previously, he has raised concerns over the connection between Kevin Shuvalov, the Austin consultant handling Dee Margo’s campaign, and Karl Rove, the political strategist credited with putting George Bush in the White House. Shapleigh has predicted a negative campaign.
Shuvalov said “we don’t talk about campaign strategy,” and would not confirm if the call came from his campaign.
The caller told me she worked for Promark. The company’s Website describes the company like this: “Promark Research Corporation, headquartered in Houston, provides state-of-the-art research, data collection and tabulation services to the research community including market research firms, political analysts, consultants, corporations, public relations, investor relations and advertising agencies, universities and other independent research organizations.” [promark website]
A Google search for the keywords “promark push poll” brings up a number of questionable polls involving the company and political races.
Still, according to the strict definition as given by the professionals, I apparently had been message-tested, not push-polled.
I should have just watched the game.
* * *
Sito Negron may be reached at sito@newspapertree.com.