Man Without Qualities

Thursday, January 15, 2004

Iowa Turmoil

Iowa is in turmoil. Well, not Iowa exactly, but the media coverage of Democratic caucus campaigning there. We are told that the race is now a statistical dead heat among Kerry (21.6%), Dean and Gephardt (both at 20.9%) and North Carolina Sen. John Edwards (17.1%).

Maybe it's actually true. But before reading too much into these Iowa tea leaves (alfalfa leaves?), it's worth noting that the turmoil centers on just one rather problematic axis: John Zogby and his polls. The poll results are Zogby Poll results. And John Zogby personally has been using his Iowa polling to garner a lot of media coverage for himself and his business. it is John Zogby who obligingly tells the news-hungry Reuters reporter John Whitesides: "We might see these candidates exchanging leads all the way to the end." No doubt that news just made Mr. Whitesides' day.

Now John Zogby is a respectable, serious pollster. He attracted considerable attention with his apparent ability to detect that sentiment moved in favor of Al Gore in the waning days of the 2000 presidential election. But there are also some other rather pertinent aspects of his modus operandi that bear keeping in mind in connection with current Iowa kerfluffle:

First, Zogby is putting very little of his credibility on the line in Iowa. These poll results have unusually high margins of error - 4.5%. That means that even accepting the premises of the current poll Howard Dean could be attracting as much as 25.4% of likely caucus attendees, with John Edwards (for example) at 12.6%. That doesn't look so much like a dead heat, and wouldn't get the coverage. But if it happened, Mr. Zogby would have a nice way out.

Worse, according to the Reuters article posted on the Zogby website, the Zogby polls include respondents who say they are likely to attend the caucuses. In evaluating who is a "likely voter" in a general election, many pollsters do not just rely on the respondents' self characterization. They also evaluate factors such as actual past voting record and the like.

In addition, it's been particularly cold in Iowa. As the Reuters article delicately phrases another issue this raises: Polling in Iowa is complicated by the unique nature of the caucus system, which requires participants to leave their homes on a typically bitter cold night and gather with neighbors for hours before publicly declaring their support for a candidate. Some reports from Iowa suggest that there is greater uncertainty as to who will turn out than is normal even in Iowa.

Although his polls and quotes are receiving a lot of media coverage and may be affecting the course of voting, Mr. Zogby's sound bites tendered to eager media representatives misleadingly fail to acknowledge just how problematic Iowa caucus polling is:

"The polls leading up to the caucuses often do not reflect the outcome of the caucuses," said David Goldford, head of the political science department at Drake University in Des Moines.

In 2000, for example, Vice President Al Gore led former Sen. Bill Bradley of New Jersey by only 4 percentage points in the final polling. In the end, though, Gore crushed Bradley, 63.4 percent to [34.9] percent snuffing out the momentum Bradley was hoping to gain in Iowa by exceeding expectations.

But more importantly as far as Zogby reliability is concerned, the very volatility Zogby is reporting itself gives him cover almost regardless of the actual caucus results. After all, where - as Mr. Zogby has noted - individual candidates can move by more that four or five percentage point in a single day, who can blame a pollster if the final swing yields results seemingly at variance with the polls? Let's see, poll margins of error of 4.5%, with 5% candidate swings in a single day? What's the predictive value of that? Not much. But the media grandstanding potential is huge!

Mr. Zogby seems to have a history of exploiting such weaknesses in polling to his own advantage. One savvy observer recently pointed out to the Man Without Qualities that if you go back and look at Zogby's numbers from 2000 and 2002, you see swings in state and national electorates that are simply beyond belief. Some media professionals politely characterize this as a "quality control" issue. There are others who will tell you that, in a pinch, John Zogby makes up the numbers. I have no evidence of that - but it has been said.

John Zogby does not just take polls - he provides unusually substantial analysis, and shares dollops of that analysis with the media. I am told by some whose views I respect that the quality of his analysis is suspect by many professionals. Oddly, this is partly why he's a media favorite: Mr. Zogby has a history of divining big swings and shifts out of high margin of error data - as he is now doing in Iowa. He gives reporters what they want (news coverage changes XYZ race in dramatic fashion) even if the evidence is a bit thin...shall we say.

I am not saying John Zogby is wrong. But I am saying he's getting quite a lot of media coverage out of polling that offers a lot less than meets the eye.

And all that's without giving consideration to the fact that John Zogby's brother is Dr. James J. Zogby, Democratic activist, former aide to Al Gore, and founder and president of the Arab American Institute, a sometimes controversial Washington, D.C.-based organization which purports to serve portions the Arab American community and which some describe as all but a branch of the Democratic Party. I have no evidence that John is influenced in his polling and analysis by James. But their close relationship does raise questions in a race in which American relations with parts of the Arab world are more than a marginal consideration.

MORE: Kausfiles has lots more, including this cite to Josh Marshall.

STILL MORE: The Man Without Qualities reprises a question on whether the New York Times and at least a good portion of the Democrat-alligned mainstream media generally don't like Dr. Dean and are skewing news coverage to deliberately undercut him. This time my question and confidence is buttressed by observations of the redoubtable Peggy Noonan:

The press has kicked in and is playing a part in the drama. The journalistic establishment has become an anti-Dean mover. Tuesday's New York Times piece on the absent Mrs. Dean, for instance--that was a piece with a sting. They decided to front-page it six days before the caucuses. The morning network news shows and the cable news shows are full of Mr. Dean's gaffes, Mr. Gephardt's rise and Mr. Edwards's potential....

Reading between the lines and listening between the lines, it's hard to avoid the thought that reporters don't really like Mr. Dean. The last time a viable Democrat rose, in 1992, the columnists for the newsmagazines and profile writers for the newspapers loved Bill Clinton with a throbbing love. None of those columns are being written now. They don't love Mr. Dean.

This is not a shock. He seems as unlovable (unless you're a Deaniac) as he is improbable. But I suspect there's something else at work. I wonder if mainstream media aren't trying to save the Democratic Party from Mr. Dean.

If Ms. Noonan is right as she so often is, John Zogby is giving these media exactly what they want to hear. Isn't that generous of him? And they, in turn, are helping to make his predictions come true.

Why it could be a virtual ecosystem!

Comments: Post a Comment