Ben Smith, who was all over the story of those ugly calls into Iowa last week (calls that included a nasty negative involving Elizabeth Edwards' battle with cancer), has posted some new details that get us much closer to the real story:
I spoke this morning to a source closely familiar with the details of that Iowa poll that raised eyebrows last week, though he was unable to definitively say who sponsored the poll, and spoke on the condition of anonymity.
The poll drew public attention after bloggers reported receiving a call asking about John Edwards' decision not to stay home with his ailing wife.
My source says the poll actually took two tracks, and included questions about all three leading Democrats.
The first question asked for the voter's preference, and then asked "what concerns you most" about two candidates.
If the voter favored Obama, the poll tested two negatives: that he "lacks the experience necessary to be president" or that he "would be a weak general election candidate"? The poll then moved on to ask about Edwards.
If the voter favored any other candidate, the pollster asked questions about Clinton, and then about Edwards.
The Clinton negatives were whether she's "insincere and changes her mind too often;" whether she "would be a weak general election candidate;" and whether she wouldn't "bring real change to Washington," the source said.
The pollster actually changed the Edwards question after complaints became public, a few days into the survey. It began by testing the message about Elizabeth Edwards' health. But after that stirred controversy, the question was changed to focus on his past as a trial lawyer.
Smith also adds that another source says "the calls originated with the Ohio-based firm Influent, which runs call centers."
Those new details are consistent with several comments left on my post last week. A commenter named Mike reported:
My wife (not a blogger) received a call like this last night. They'd been trying to reach us since Monday. Caller ID listed all zeros. When asked, She listed her first choice as Obama, and was subsequently asked the negative questions about Obama and Edwards.
Another reader that went by the name "canary" reported:
The call that I received was a bit different: after asking for whom I would caucus and then my second choice, I was asked: Of these alternatives, which is most important to you: a) that the nominee have sufficient experience to run the country, or b) that they be electable in the general election.
The next question was:
In terms of electability, which of these factors would cause your concern about the nominee: a) that they are regarded as too liberal, or b) that the candidate is a trial lawyer.
I followed up via email with the last commenter, who says she received the call on Saturday night and confirms that the questions she heard did not attach specific names to the candidates: "it was left to me to draw a conclusion as to whom the caller was referring."
One more bit of information worth promoting from the comments. As suggested by reader Willie, anyone calling from a list of registered voters available from the Iowa Secretary of State would have had the basic demographics (age, gender and geographic location) as well as enough non-caucus voting history to do basic "likely voter" selection. Some legitimate micro-targeters do just that in order to create a very short interview (devoid of demographic questions) that allows for thousands of calls in as little time as possible.
Smith's reporting and the comments above help provide a much clearer picture of the motives behind these calls. Influent is a firm that does telemarketing, not survey research calls. The survey involved just five or six questions, but included questions about strength of support and second choice -- two important measures if your aim is to identify very small target subgroups ("micro targets") from a sample of thousands of potential Iowa Caucus-goers. Moreover, the fact that the sponsors switched the wording more than once, first to a less offensive "negative" about Edwards and then to a version that did not attach specific candidate names to the negatives, all suggests this was some sort of micro-targeting or "data harvesting" project.
I remain convinced that the odd structure and language of the questions is not the work of an experienced campaign pollster. However, I am now doubtful that the sole motive was to spread a nasty negative about John Edwards. The initial questions asked, the call center used and the fact that that the sponsors apparently changed the "negatives" more than once in reaction to the public backlash, all suggest that the main purpose of the project was some sort of micro-targeting. However, the pattern also suggests the sponsors were indifferent, at best, to including an explosive personal attack on John and Elizabeth Edwards, a fact they will have to answer to if their identity is revealed.
Ben Smith reminds us that their identify "remains a mystery," and urges readers to contact him with any relevant details.
Update: Ben Smith's latest update quotes Ted Bernard, senior vice president at Influent, Inc, who, saying he can "neither confirm or deny" that the calls originated at his facility.
Follow Mark Blumenthal on Twitter: www.twitter.com/MysteryPollster