BOB GARFIELD: This is On the Media. I'm Bob Garfield.

BROOKE GLADSTONE: And I'm Brooke Gladstone. Historians, looking back at October of '04, might remember it as the month of the sweep. Not only the four games to zero triumph of the Boston Red Sox, but also the three for three performance of John Kerry in the presidential debates. Those historians need only glance at the headlines. A new study finds that in the first two weeks of October, more than half of the mainstream news stories about candidate Bush, and not just about the debates, were negative in tone, compared to only a quarter of candidate Kerry's. And the study noted that this was a mirror image of the closing weeks of the last presidential election, when Bush received twice as much positive coverage as Al Gore. Tom Rosenstiel is the director of the Project for Excellence in Journalism which conducted the study, and he joins me now. Tom, welcome back.

TOM ROSENSTIEL: My pleasure.

BROOKE GLADSTONE: So, first of all, how did you determine what was "negative" as opposed to what was positive press coverage?

TOM ROSENSTIEL: For a story to be clearly negative, the critical or unfavorable assertions in that story had to outnumber the positive assertions about a candidate in that story by two to one. Also, the stories had to be about a candidate -- it couldn't be a story about, you know, the electoral process or--

BROOKE GLADSTONE: Right.

TOM ROSENSTIEL: -- you know, voter habits. These were stories that were focused on a particular candidate or were comparative about the two candidates together.

BROOKE GLADSTONE: Now when you talk about "assertions," do you mean statements of fact?

TOM ROSENSTIEL: No. What we mean is either quotations that are from sources, or the narrative from the author of the piece that are making judgments such as-- "White House was scrambling to put the pieces together all day today, after a poor debate performance last night."

BROOKE GLADSTONE: Mmm-hmmm.

TOM ROSENSTIEL: That's an assertion that has some interpretative or judgmental sense of it.

BROOKE GLADSTONE: Let's take your example: "The White House is scrambling after the president's poor performance in the debate." Now the polls suggested that the president gave a poor performance in the debate. Is that, then, interpretative or is that a statement of fact?

TOM ROSENSTIEL: Well, if the statement that polls showed that the majority of people didn't believe that the president won the debate, that would be a statement of fact. But the story I'm thinking of said "President Bush signed a tax bill into law today, bah, bah, bah, bah, but behind the scenes, the White House was scrambling to recover from a poor debate performance." So those were choices that the reporter made, and, and they were not descriptive of events, they were actually descriptive of a, sort of a context in which he was trying to put the events.

BROOKE GLADSTONE: I wonder why you frame your study in terms of negative and positive coverage. Is it possible that these categories obscure the bigger picture about what's really important in reporting -- whether it's true -- whether it's contextualized? Negative versus positive may not be very meaningful.

TOM ROSENSTIEL: Well, that's only one of about eight variables that the study looks at. It's something, however, that there's a growing interest in, and, and it's easy to do, and I think more relevant to do in terms of, of a debate period, because you do want to kind of get a sense of who prevailed in the coverage around this time, and frankly, I think the fact that Gore's coverage and Bush's coverage are such exact mirror images of each other, tell us something surprising and, and interesting which is that what happened to Gore four years ago happened to Bush this time. And it changed the race four years ago, and it's going to change the race this time.

BROOKE GLADSTONE: But last time around, Gore was called a liar for inconsequential exaggerations.

TOM ROSENSTIEL: Mmm-hmmm.

BROOKE GLADSTONE: He was called a liar for things he didn't actually say, such as being the model for the protagonist in Love Story and so on. These were unfounded snarky observations that were made by the media all the time. I mean it may add up to the same negativity numbers, but it doesn't mean the same thing happened.

TOM ROSENSTIEL: Brooke, what we're looking at now is just the debate phase, and the fates of these two candidates who were perceived to have gotten the worst of the debates. That's where the parallels and behavior are. But, you know, I have to tell you that conservatives think that the press has it out for the president, just the way that a lot of Democrats thought the press had it out for Gore four years ago.

BROOKE GLADSTONE: I guess I'd just really like to see a study done every four years about which candidate is being treated fairly and which candidate is being treated unfairly, rather than negatively and positively.

TOM ROSENSTIEL: Yeah. I mean that's what you're ultimately you want to try and get a sense of. The problem is what's fair is so hard to get your hands on that it becomes a cumulative judgment. Tone becomes a component of fairness, but it doesn't get you the full picture.

BROOKE GLADSTONE: Why do you think we're seeing more negative coverage of Bush now?

TOM ROSENSTIEL: In an environment where the press is often being accused of bias, and when the race is confusing and mixed up, the debates -- and particularly these debates -- where clearly Kerry performed better -- were a moment when everybody could agree on something and there was a sort of rush to a consensus. I think another reason is that the press coverage in general has become more interpretative over the years. Reporters, particularly covering politics, feel freer to, you know, interpret and even make judgments in their stories. Only 14 percent of the stories in this year's study were straight news, that old inverted pyramid style, whereas four years ago, we saw in the same time period twice as many, 29 percent of the stories, were the sort of straight news, just describing the events of the day. We're not suggesting that this is appropriate or inappropriate.

BROOKE GLADSTONE: I got a clear sense from the remarks that accompanied the report that you didn't think interpretive reporting was very good.

TOM ROSENSTIEL: Well, I, I know why editors believe that their, their coverage needs to be more interpretive than it used to be. In the age of 24 hour news and the internet and satellites and cable, facts, mere facts, are considered common commodity, and your newspaper the next morning, your evening newscast that night has to do something more than recite the facts that people may already know. But I also believe that in an atmosphere where charges of bias are a lot more rampant than they used to be, you know, these things are not so simple. Interpretive reporting is in some ways a higher form of reporting, but it's also a riskier form of reporting.

BROOKE GLADSTONE: Tom, thank you very much.

TOM ROSENSTIEL: My pleasure.

BROOKE GLADSTONE: Tom Rosenstiel directs the Project for Excellence in Journalism which has just released a new report on campaign coverage in the first two weeks of October.