Articles and Analysis


Pro-Clinton Push Poll?

Topics: 2008 , Hillary Clinton , John Edwards , LA Times , Pollster , Push "Polls" , The 2008 Race

Andrew Malcolm of the Los Angeles Times reports evidence of a pro-Clinton "push poll" in California, or as he defines it, "malicious political virus that is designed not to elicit answers but to spread positive information about one candidate and negative information about all others under the guise of an honest poll."

His definition is right, but does the call in question meet it? Malcolm's source, a former local television news director named Ed Coghlan, describes a call from "a pollster who wanted to ask registered independents like Coghlan a few questions about the presidential race." The survey tested reactions to statements about Hillary Clinton and negative statements about Barack Obama, John Edwards and John McCain:

Coghlan said he was offended by such underhanded tactics and knew he was going to get out a warning about this dirty trick, but he said he played along for the full 20-minute "poll."

That last bit of information tells me that this call was almost certainly a message testing survey, and not a so-called "push poll." California has over 15 million registered voters, and roughly three million of those are independents. If "someone" was paying "to spread this material phone call by phone call among independent voters," would they really spend 20 minutes on the telephone with each one?

Not likely.

The call that Coghlan describes sounds more like a message testing survey that included many negative messages about Clinton's opponents. In other words, someone called a random sample of voters with the intent to "elicit answers," or more specifically reactions, to negative messages that the Clinton campaign or an allied group considered airing in California.

Negative campaign messages may be offensive, unfair or untrue, and it would certainly be reasonable to question the Clinton campaign on the fairness or truthfulness of the messages tested in this call. Legitimate message testing surveys sometimes cross ethical lines, especially when they raise explosive topics that candidates are unwilling to discuss openly (see the controversy over the calls in New Hampshire and Iowa that tested negative messages about the Mitt Romney's religion).

In this case, however, the only specific negative message that Coughlin reports is the attack on Barack Obama for his "present" votes in the Illinois legislature. Both Clinton and Edwards raised that issue openly in the South Carolina debate.

So far, at least, Malcolm's claim to have uncovered a "malicious political virus" operating "under the guise of an honest poll" is not supported by the facts reported.

For further reading: We have discussed the distinction between so-called "push polls" and message testing many times. Most relevant are my comments on the distinction between "push polls" and message testing here and here a well as those by Stu Rothenberg, Republican pollster Neil Newhouse and the statement from the American Association for Public Opinion Research (AAPOR).


Michael McDonald:

New national CBS poll has Clinton & Obama tied at 41.


Dr. Brad Burke:

Either way, it's underhanded and typical of the very politics we must move beyond. Apparently, the Clinton camp didn't understand the "Change" memo.

P.S.- For the Clinton camp, change means a fundamental shift in the way we approach politics in America, not simply getting rid of Bush.


Mark is of course right: it's a message test of 500 to 1000 voters, not a push poll. Sadly, the denial -- or correction -- never keeps pace with the report. Something similar happened to us once, when a reporter got a call from the mom of a volunteer for an opposing campaign, furious that she had been called for a message test poll that aired criticisms of her son's candidate to 800 voters. She called a newspaper ... and with the combination of ignorance about polling and the "gotcha" mentality of today's press we were soon looking at a headline accusing us of push polling. Had the reporter in that case -- or this -- done the most elementary homework it would have been clear this was not the case. When a true "push poll" is launched -- short calls with negative (and usually false) information to tens or hundreds of thousands of voters -- it does not stay quiet for long, because news media switchboards light up like Christmas trees, as they did in South Carolina recently.


Richard Berman:

I'm not a pollster, but I have a question for those who are more informed. Assuming the poll in question actually wants to test positive questions about one candidate, in this case Clinton, and negative about others, in this case Obama, Edwards and McCain, then what is the difference in the effect of this poll and the effect of a push poll? In other words, the caller who heard these questions would tend to become positively disposed to one candidate and negatively disposed to the others. Even if the intent is not to conduct a push poll, isn't the effect the same?


Andrew Long:

Another fair question I think: the call is reported as having happened "the other night," which may mean any day last week. Since Clinton was up on the air already in CA with at least two different ads, and had dropped all sorts of mailers and other advertising, what would a "message testing" call like this hope to learn at this late date? And wouldn't this kind of effort have been concluded weeks ago?



Yeah, I was going to ask Andrew's question... why message test this week?

There a couple of important differences:
1) You can't spread a message by calling 1000 voters (especially in a state the size of California).
2) You aren't lying about the reason for your call -- you're actually asking questions for their research value.


Great site!

Would you like a Link Exchange with The Internet Radio Network? At the IRN you can listen to over 60 of America’s top Talk Shows via Free Streaming Audio! In addition you can email the President, VP and Congressional Leaders!!



I received the same kind of call here in NH before the primary from the Clinton people. It was one anti-Obama argument after another. It seems to me that it has the same effect as a push poll. Yes fewer people are called but the "poll" seeks out the most effective negative argument to be broadcast more widely through other means. I don't see how this kind of polling is in any way ethically superior to a true push poll. After listening to Obama get repeatedly attacked in the context of various "questions" I told the pollster, "I am not going to help Hillary Clinton find the best way to trash Obama." Then I hung up.



@Mark, Craig:

I feel confident agreeing that your assertions are correct -- that this is denotatively not a push-poll. However, and as Stephen ponts out, it doesn't matter. What matter is that this method has a polarizing effect on its audience. Even as an aredent Bill Clinton supporter during the 90's, I think we can all be pretty sure that this is precisely the intention. Oh, those glorious ends! Means, right this way...

But even if my suspicion be incorrect, it says something that my initial reaction is blame-oriented. The Clinton 08 campaign has thusfar demonstrated more than amply that they are not to be trusted under any circumstances, so quick and unapologetic is their instinct to get dirty.

So perhaps we ought to put this kind of questionable message-questioning in a category of its own. I believe Nixonians had a name for it, something beginning with "rat."



If the most negative message in a 20-minute call was Obama's 'present' votes, it doesn't seem very malicious. Or worth 20 minutes of phone time for the sponsors, whoever they are -- unless they really are researching messages for use in ads. And if that's the most negative ad message they're considering -- then they deserve a medal.:-)


Post a comment

Please be patient while your comment posts - sometimes it takes a minute or two. To check your comment, please wait 60 seconds and click your browser's refresh button. Note that comments with three or more hyperlinks will be held for approval.