Howling Dog Graphic
Point. Click. Search.

Contents: Archives:

Search this weblog
Search WWW
Howler Graphic
by Bob Somerby
E-mail This Page
Socrates Reads Graphic
A companion site.

Site maintained by Allegro Web Communications, comments to Marc.

Howler title Graphic
Caveat lector

29 September 1999

Our current howler (part II): Bad with names

Synopsis: A recent Pew survey revealed a fact the press corps rarely mentions.

Commentary by Chris Matthews, Tom Squitieri
Hardball, CNBC, 9/15/99

Commentary by Tony Snow
Fox News Sunday, Fox, 9/19/99

Commentary by Alexis Simendinger
Hardball, CNBC, 9/16/99

Guess Who’s Running Now?
Howard Kurtz, The Washington Post, 9/16/99

What's wrong with letting scribes play around with concepts like "statistically even?" Some of the pundits start cutting corners—and start telling the story they like. Here, for example, was a tabloid talker at the start of his September 15 program:

MATTHEWS: What is going on out there? I thought Al Gore had the thing made. I thought he'd get a decent challenge by Dollar Bill Bradley, but now Bradley is tied in the latest New York state poll. He's tied with him in the latest New Hampshire polls. These polls are matching up. They're confirming each other. Something is going wrong. Al Gore has six pollsters working for him in New Hampshire trying to change the numbers. But people don't like the dog food apparently.

But if Gore is trying to "change the numbers," he should hire the tabloid talker. The New Hampshire polls which the talker called "tied" all showed Gore ahead (by four, five, and seven points; see THE DAILY HOWLER, 9/28/99). In the New York poll which the talker called "tied," Gore had a two-point lead. Where others had spoken muddily about "statistical" ties, the talker had dropped the qualifier entirely. Four days later, our buddy Tony Snow got carried away, and somehow had Bradley ahead:

SNOW: Let's switch now to presidential politics. We've seen all the Washington papers, anyway, New York papers are filled with stories, Mara, that Bill Bradley and Al Gore—

MARA LIASSON: That's right—

SNOW: —are even, or Bill Bradley even edging ahead in New York and New Hampshire.

As we saw yesterday, this was the morning when a Post page-one headline said Bradley had pulled even in New Hampshire, although the article cited only the three statewide polls in which Gore actually led. The headline writer had inaccurately called Bradley "even;" on that basis, Snow put him ahead!

Yep. It's just a hint of the mischief that can ensue when scribes play around with "statistical" ties. At the time, with three polls all showing Gore ahead, there was little chance that the race was actually tied, and no reason at all to think that Bradley was somehow actually leading.

After his September 15 fumble, a tabloid talker took two nights off. But the next night, with guest host Lawrence O'Donnell discussing more polls, an irregular guest threw in a statistic that raised an intriguing question:

SIMENDINGER: I think one of the things we should talk about a little bit is a surprising outcome in a Pew poll this week, a poll of 1200 Americans in which 37%, given the opportunity to volunteer a name of someone running in the Republican party, could not name a single individual, not Bush or anybody else. 50% could not name a Democrat who was running.

Howard Kurtz had discussed the poll in that morning's Washington Post. In fact, 54% of Pew respondents couldn't name a Democrat running for president. Kurtz called the result a "stunning finding"—a measure of how little attention Americans are paying to the early stages of the 2000 race.

Why had Simendinger mentioned the survey? Here is what she said:

SIMENDINGER: When you look at numbers like that, you ask yourself what's going on with some of the specific numbers we're looking at, and I think there's a lot of truth in this idea that American voters at this point are not focused in on this, they don't know a lot about the candidates, and they're reacting, I think, somewhat superficially to what the news may be week to week, day to day, or to the overall trend that they're reading or hearing about.

Or to what the spin may be night to night. In fact, Simendinger was being polite when she said the voters didn't know "a lot" about the candidates. The truth is, half of them don't even know the candidates' names, which does in fact raise an obvious question about our discussions of current polls. If (as Kurtz reported) only 54% of Americans know Bush is a candidate, and only 46% of Americans know Gore is a candidate, then how much meaning can one invest in the result of a preference poll? (Only 16% knew Bradley was running.) What exactly can we glean from preference polls, when half of Americans don't even know that Bush and Gore are running for office? The question didn't seem to interest O'Donnell; when Simendinger raised the matter, he immediately moved on, asking Jerry Nachman to review recent Bradley stump speeches. But here at DAILY HOWLER World Headquarters, the analysts roared in our vast viewing chamber, lustily cheering Simendinger's vain effort to deepen the tired debate.

Did the Pew report offer a "stunning" finding? Clearly, in one way, it did. By any standard, it is remarkable to learn how uninformed the electorate actually is. The finding flies in the face of standard presumptions about the way our democracy functions. As Kurtz noted in his column, the Pew poll, conducted from September 1-12, followed several weeks of intense media coverage of the silly Bush cocaine flap. Even after that exciting coverage, 46% of Americans still didn't know that Bush was running for president.

But in one other way, the report was hardly surprising, because information surveys routinely reveal the public's unshakable ignorance of almost all hard news topics. What made the latest example seem "stunning?" Only the fact that the media rarely reports this state of affairs. In the current instance, for example, O'Donnell wasn't the only news leader uninterested in the Pew results. In the five major papers we cover, only Kurtz discussed the Pew findings. Other papers plowed ahead with their normal discussions of polls, never telling their readers a significant fact—that the respondents cited in the polls don't have a clue about the unfolding election.

When the press discusses poll results, it rarely reveals an important fact—respondents routinely lack factual knowledge of issues on which they're opining. Another remarkable recent survey explored this striking state of affairs.


Tomorrow: It's the law: when the public's ignorance does come to light, the press offers pleasant excuses.

Just as we said it was: James Dao, in this morning's New York Times, on Senator Bradley's newly-revealed health plan:

DAO: The former Senator said his plan would not require new taxes, and could be financed using the non-Social Security surplus, projected to be more than $1 trillion over the next decade, as well as cost-saving measures. But he underscored the preliminary nature of the proposal by saying he did not have a backup financing plan if the economy turned sour and the surplus evaporated.

God help us all—where to begin? Dao's own newspaper has carefully explained that the "projected surplus" is based on spending caps that no one believes will be kept (see THE DAILY HOWLER, 8/23/99). The projected surplus will "evaporate" even if the economy stays healthy. This fact has been explained again and again, most recently in a Nancy Gibb lead story in Time ("Phantom Surplus," 9/20/99). But, because both the White House and the Congress are pretending the surplus will exist, the daily press corps is obediently playing along too (see THE DAILY HOWLER, 8/27/99). Yesterday, Bradley became the latest major player to spend the phantom future money, without anyone is the press corps asking him if he thinks the future caps will really be kept. Would Bradley keep the future caps? Surely he wouldn't—but no one asked him. Again we see the press corps' instinct for proceeding in obvious error.