DEFINING EXCELLENCE DOWN! Did Democrats get “more favorable coverage?” We don’t have the slightest idea: // link // print // previous // next //

YOUR CHANCE TO WATCH HISTORY IN ACTION: In this post, Digby records Chris Matthews insulting Clinton on Sunday’s Chris Matthews Show. (“A conceited goody two-shoes.” He also vastly dissembled.) In this post, Media Matters records the same guy insulting Clinton on last night’s Hardball. (Decent people who endorse Clinton are “castratos in the eunuch chorus.”)

As we’ve told you, Matthews went into full propaganda mode in the weeks leading up to that crucial October 30 debate; he has been staging a pure propaganda campaign from that time to the present. Of course, this is exactly what he did, over a longer period, in the two years of Campaign 2000, when he endlessly trashed—and dissembled about—that ludicrous Candidate Gore. (“A man-like object” who “would lick the bathroom floor to be president. The bath-tub ring.”)

You’re seeing a type of history here. Whomever you prefer for the Dem nomination, this is your chance to see pure propaganda being delivered under the guise of “cable news.” Matthews is profoundly dishonest—and this is your chance to see it played out, as he shapes world history again. And this is your chance to see the press corps’ stooges agree with him nightly. On Sunday, Matthews dragged out the leering chimpmunk Andrew Sullivan (one of Josh’s absolute favorites) to trash and bash the current Dem leader (though just barely). Indeed, two of Matthews’ four enablers had been brought in from other countries that day! Talk about immigration problems! Now we even import the scribes who assault us about our elections!

Whomever you prefer for the Dem nomination, this is something you should make time to see. Gather the kids, and make them watch too. They’ll be seeing the way a wealthy elite (disguised as a “press corps”) takes control of a White House election. Eight years ago, most liberals didn’t know that this was going on. Here’s your shot at the reruns.

For ourselves, we’ve come to believe that this mountain of narrative would vastly affect a Clinton White House. (Trust us: These fools would have driven Gore into the sea.) Since it’s clear that liberals and Dems will never have “what it takes” to resist this misconduct, we think this conduct must be considered in casting a primary vote.

But whatever you think about primary voting, don’t miss the demolition derby being conducted under the guise of “cable news.” This conduct sent George Bush to the White House—while liberal leaders kept very quiet. Thanks to that silence, Jack Welch’s Lost Boy is making history again. Notes on current silence tomorrow.

DEFINING EXCELLENCE DOWN: Who would be more effective as president—Edwards or Obama? We feel less sure than Paul Krugman does. We were also surprised to see The Man offer these views in yesterday’s column:

KRUGMAN (12/17/07): There's a strong populist tide running in America right now. For example, a recent Democracy Corps survey of voter discontent found that the most commonly chosen phrase explaining what's wrong with the country was ''Big businesses get whatever they want in Washington.''

And there's every reason to believe that the Democrats can win big next year if they run with that populist tide. The latest evidence came from focus groups run by both Fox News and CNN during last week's Democratic debate: both declared Mr. Edwards the clear winner.
[...]

So what happens if Mr. Obama is the nominee?

He will probably win—but not as big as a candidate who ran on a more populist platform.

Would “a candidate who ran on a more populist platform” do better in November than Obama? We don’t have the slightest idea—and we don’t know why Kruggers believes that he does. But we were struck by something else Krugman said in this column. We were struck by his reference to a survey which has now gained a fair amount of attention—more than it deserves, we would guess.

That survey was conducted by Pew’s inaptly-named Project for Excellence in Journalism. Let’s quote Krugman: “According to a recent survey by the Project for Excellence in Journalism, Mr. Obama's coverage has been far more favorable than that of any other candidate.” To be more precise, the survey was released fairly recently (on October 29), but it deals with press coverage in the first five months of this year, from January 1 through May 31. And it’s true: The survey did find that “Obama's coverage has been far more favorable than that of any other candidate.” But how much confidence should we place in this survey? Not that much, we’d have to suggest.

In our view, Obama has almost surely gotten better coverage than Clinton or Edwards. But the cited survey is very shaky. Since it’s being cited fairly often now, it might be worth seeing why.

As noted, the survey was released on October. On this, its “Overview” page, it makes some basic assertions:

PROJECT FOR EXCELLENCE IN JOURNALISM (10/29/07): In the early months of the 2008 presidential campaign, the media had already winnowed the race to mostly five candidates and offered Americans relatively little information about their records or what they would do if elected, according to a comprehensive new study of the election coverage across the media.

The press also gave some candidates measurably more favorable coverage than others. Democrat Barack Obama, the junior Senator from Illinois, enjoyed by far the most positive treatment of the major candidates during the first five months of the year—followed closely by Fred Thompson, the actor who at the time was only considering running. Arizona Senator John McCain received the most negative coverage—much worse than his main GOP rivals.

According to the study, Obama had enjoyed “by far the most positive treatment” from “the media” among the five front-running candidates. (Put another way, he was getting “more favorable coverage” than the other candidates. A chart is provided.) In short, the study purports to measure who is getting “favorable coverage” or “positive treatment”—from something described as “the media.” But how did the study attempt to accomplish this task? In our view, it’s all downhill from here.

Who is the media? For openers, this study seems to follow a shaky procedure as it defines who “the media” is. Just consider how it decided which newspapers to sample. On its “Methodology” page, the projects lists thirteen newspapers which were sampled—but makes little attempt to explain how they were selected. And we can’t help noting a certain tilt in some of the selections. The New York Times is sampled, for example—but not the conservative New York Post. Ditto for the Washington Post and the Boston Globe; both are sampled, but their conservative counterparts, the Washington Times and the Boston Herald, are not. In short, we would say that this project tends to exclude more conservative newspapers—and there seems to be no real rhyme-or-reason to the way these thirteen papers were picked.

Do these thirteen papers represent a faithful sample of the nation’s newspapers? We can think of no reason to assume that; essentially, the project seems to have picked-and-chosen, largely at will. Meanwhile, the project ends up declaring that Democratic candidates got more “favorable coverage” during this period than Republicans did. Could that finding in part reflect the apparent exclusion of conservative newspapers? We have no way of knowing. But this seems like a very weak research model—a standard component of major projects done by this inaptly-named crew.

By the way, only front-page stories were reviewed. (Beyond that, for unknown reasons, Saturday stories didn’t count.) In short, all op-ed columns and editorials were skipped; so was all inside-the-paper coverage. That is a perfectly valid research strategy, but it vastly limited what got considered by this project. For example, all that trashing of Democratic wives (and haircuts) by Maureen Dowd ended up on the cutting-room floor. In this way, the study excludes a large part of what people think of when they think of “the media.” Over the five-month period in question, only 168 stories from these thirteen newspapers ended up in the study.

We see problems in the way other parts of “the media” were sampled. But the selection and use of the newspapers seemed especially problematic.

What is favorable coverage: A second point is even more problematic. What does this study mean when it talks about “favorable coverage” or “positive treatment?” It’s very hard to answer that question—and to the extent that it can be answered, it seems that the Project may not be measuring what the average person thinks of when he hears such phrases.

What does the Project mean by “favorable coverage?” On this page (entitled “Tone”), we get this first explanation:

PROJECT FOR EXCELLENCE: The volume of coverage is one thing. But in politics, not all coverage is equal, even if they spell your name right. What was the tone of the coverage each candidate received?

While Hillary Clinton may have gotten the most press, she did not get the most favorable. That distinction, among major candidates, went to Barack Obama.

[...]

To evaluate tone, the study examined every assertion that offered some assessment of a candidate’s chances at winning or their potential effectiveness in office if they were elected and tallied them by story. For a story to be considered positive or negative, two thirds of all the assertions had to be explicitly positive or negative in tone or the story would be considered balanced.

It’s hard to know just what that means—and the project seems to include no samples of newspaper stories judged to be “positive” or “negative.” But for what it’s worth, a wide range of favorable and unfavorable comment would to seem to fall outside the rather narrow guidelines described in that passage. Would a reference to John Edwards’ troubling haircut (or to his troubling house) count as an “assertion that offered some assessment of a candidate’s chances at winning or their potential effectiveness in office if they were elected?” Taken literally, no—it wouldn’t seem to. And yet, this sort of thing is what most Democrats think of when they think about “negative coverage.”

At any rate, that explanation leaves us uncertain; just what statements were included in this study? But on the “Methodology” page, a second explanation is offered. And uh-oh! Unless we’re missing something here, it doesn’t seem to say the same thing as the first explanation. In this passage, the Project once again seems to explain how it measured “tone:”

PROJECT FOR EXCELLENCE: Primary figure tone reflects whether the journalist’s tone is constructed in a way, via use of quotes, assertions, or innuendo, which results in positive, neutral, or negative coverage for the story’s primary figure as it relates to the topic of the story. While reading or listening to a story, coders tally up all the comments that have either a negative or positive tone to the reporting. Direct and indirect quotes are counted. In order for a story either positive or negative, it must have 1.5 times the amount of positive or negative comments (with an exception for 2 to 3, which is coded as neutral). If the headline or lead has a positive or negative tone, it should be counted twice into the total value. Also counted twice for tone are the first three paragraphs or first four sentences, whichever comes first.

In that explanation, it seems that a much wider array of statements would be part of the measurement. (Edwards’ haircut and house are now in.) And then too, the mathematics seems different. In this second explanation, it seems that a story with six negative comments and four positive comments would be ranked as “negative.” Under the first explanation, it seemed that it would not—that the story would count as “balanced.” These two passages seem to explain the same thing—and yet, they seem contradictory.

Do these relatively trivial points really matter? Only if you want to cite the findings of this study. Unfortunately, we find here what we often find when we examine work by the Project for Excellence; we find sloppy, poorly-explained, contradictory thinking about the study’s most basic elements. And yes, we think this is fairly common when this group produces such work. For example, in one of its most “famous” studies, the Project examined the coverage of the Bush/Gore “character issue” in the first five months of the year 2000. On that occasion, the Project assembled a rather shaky research model—and then, the Project plainly misstated its own (rather striking) data (see THE DAILY HOWLER, 12/20/02). Unless we’re missing something here, the same kind of sloppy, inept work is found in this most recent effort.

In conclusion: Let’s conclude with two basic points.

First, the absence of sample stories seems significant. Our guess: If you saw the news reports which were scored as “positive” or “negative,” you might see that these stories do not correspond to your idea of these designations. For example: If we understand this study’s procedures correctly (based in part on prior Project studies), a perfectly objective news report will be coded as “positive treatment” or “favorable coverage” because it includes accurate statements about successes in polling and fund-raising. An example: During the period in question, Candidate McCain was polling poorly—and his fund-raising was causing problems. Did accurate reports of these basic facts lead to his designation as the guy with the “least favorable” coverage? We don’t know, but if it did, that isn’t what people think they’re learning when they hear this survey’s results.

Second, the sloppy work we seem to find here tells you about the world you live in. Within the Village, the Project for Excellence is thought of as excellent—as the gold standard of media analysis. But uh-oh! Over the years, we’ve found that its work is often quite poor. It’s the kind of work a palace elite might sponsor about its own doings.

Did Democrats get “more favorable coverage” during the first five months of the year? We don’t have the slightest idea. But then, we’ve looked at the study.