Caveat lector

Our current howler (part II): And now for what actually happened

Synopsis: The reporting on Bush was more positive everywhere. But bumbling Pew buried that fact.

Selective Negativity
Michael Kelly, The Washington Post, 5/16/01

The First 100 Days: How Bush Versus Clinton Fared in the Press
Project for Excellence in Journalism, 5/01

"How Bush Versus Clinton Fared in the Press?" We think Pew’s excellent title was written in English, although we can’t really be sure. But we can be sure of one simple fact–this study sets a daunting new standard for Misleading Nugget Assertions. Everyone knows what the Pew study says. Once again, we’ll let Mike Kelly limn it:

"Bush’s overall coverage was actually less positive than Clinton’s," the Pew report confidently states. "Contrary to Democratic complaints, George W. Bush has not gotten an easier ride from the American media" (our emphasis).

Has President Bush had an easier ride? Here at THE HOWLER, we don’t really know. But then, neither do the vastly overrated people who puzzle and ponder for Pew. The trumpets sounded all last week as they served up their latest report; the Pew people–who are willing to say sixty days is a hundred–are the press establishment’s symbol of "excellence." Yep–in this case as in so many others, their word has simply been taken as law. We all now know that President Bill got a better early ride than poor Bush.

Except we don’t know that at all, dear friends; in fact, this report found precisely the opposite. As Kelly notes, the Pew researchers studied "stories" from three separate branches of the media–TV, newspapers, and news magazines. They rated those stories as "positive," "negative" or "neutral" in tone, and then compared Bush’s coverage with that dished to Clinton. And here’s the fact that you haven’t heard–in all three areas, news reporting on President Bush was in fact more positive, according to Pew, than the news reporting on President Clinton. According to Pew, Bush had more positive reporting from all three groups–from TV, from the mags, from the papers.

And no, we kid you not. Let’s consider TV networks first. Pew studied ABC, NBC, CBS and PBS. And the findings? According to Pew, 22 percent of TV reports about Clinton were positive; 30 percent of reports were negative (the rest were judged to be neutral). Bush’s numbers were substantially better; the same 22 percent of reports were positive, but only 16 percent of reports were negative. That’s right–Bush got substantially better TV coverage than Clinton, according to Pew’s own report.

How about the news magazines? Lazily, Pew studied only one–Newsweek. But Bush also had an advantage there. Clinton’s numbers: 22 percent positive, 39 percent negative. Bush: 30 positive, 35 negative. (With Newsweek, Pew studied a blend of "stories." Some of the "stories" were news reports, some were opinion columns. No breakdown was provided.)

So Bush did better on the networks, and Bush did better in Newsweek. All that was left was the newspapers. Again, Pew selected a laughable "sample," studying only the New York Times and the Washington Post. But what did they find about the news reporting done by those two papers? Here, we encounter a bit of a problem; due to Pew’s haphazard presentation of data, the report doesn’t give the figures for the news reporting in the Times and the Post. Instead, Pew jumbles the two papers’ news reports in with their editorials and op-eds, reporting the overall data. But Pew does give enough information so that the data on news reports can be deduced–and again, Pew found a clear advantage for Bush. In the newspapers studied, Clinton’s news reports were 24 percent positive, 27 percent negative. Bush news reports, by contrast, tilted positive, 26-19. Pew does give the data for hard news stories in TV and newspapers combined; as with everything else, they favor Bush. In this measure, Clinton was 23 percent positive, 28 percent negative. Bush was 24p-18n, according to Pew’s own reported data.

So let’s go back and look at those findings. Bush got more positive reporting on TV; more positive reporting in the Post and the Times; and yes, more positive reporting in Newsweek. And those were the only outlets which the Pew people studied. How is it, then, that we keep reading that "contrary to Democratic complaints, George W. Bush has not gotten an easier ride?" That is due to the first of many methodological howlers on which Pew built its report.

Here’s what happened: as we’ve mentioned, the Pew people didn’t just study news reporting when they looked at the Post and the Times. They also looked at the editorial and op-ed pages (as is perfectly reasonable, by the way). And in the editorial and op-ed pages of the Post and the Times, they found a big advantage for Clinton over Bush. Duhhh–both these editorial boards supported the Clinton budget plan, and don’t support the Bush proposals. Unsurprisingly, the numbers came out in Clinton’s favor, by a lot; for Clinton, 37 percent of op-eds and editorials were positive, 26 percent were negative. Bush’s numbers were vastly different: 18 percent positive, 46 percent negative. And when Pew added these "stories" to the rest of the "stories" from all of the media, it produced a slight overall tilt in Bush’s favor. Overall, of the 899 "stories" Pew reviewed, Clinton ended up with a modest edge. His "stories" were 27 percent positive, 28 percent negative, while Bush was 22-28.

Surely, there’s nothing wrong with trying to measure editorial sentiment at these two papers. But two big things are groaningly wrong with the way Pew reported its effort.

First, Pew lumped the editorials in with the news reports, and described them all as "stories." Let’s ignore the kindergarten language; in lumping the editorials in with the news, Pew was mixing apples with kumquats–and in the process, Pew obscured the study’s most striking result. It’s hardly news that the Post and the Times editorial boards don’t support the Bush proposals. It is news to learn that, in news reporting, Bush has gotten more positive press in every part of the media which Pew reviewed. Lumping the editorials in with the news reports completely obscured this major finding. Indeed, what exactly has been said in those "Democratic complaints" that the Pew folks shot down? Generally, Dems have complained that Bush has gotten more favorable reporting from the mainstream press. Pew found that to be true in every area, and then–bungling and blundering as is its wont–stumbled onto a way to obscure it. (We say Pew "stumbled onto a way" to avoid implying that Pew did this on purpose.)

But there’s one other problem–groaning; howling–with the newspaper part of Pew’s study. That involves their choice of papers. Pew itself, deep in its report, refers to the Post and the Times as "liberal" newspapers. So why in the world were these two–these two only–chosen to serve as a "sample" of newspapers? Pew doesn’t say in its hapless report, but this decision makes a joke of Pew’s entire study. Stop and think: In its "newspaper" study, Pew seems to have found that, even at these two "liberal" papers, the news reporting was better for Bush! What in the world might they have found if they had looked at some other newspapers? If they had reviewed the New York Post and the Washington Times, for example? What was the tone of the news reporting at these two conservative papers? And does Pew think that the opinion writing at the Washington Times would have matched that found up in Gotham?

We hate to be so negative, folks, but his report is a total joke–the latest example of the lazy, inept work so typical of today’s press establishment. Pew found that Bush got better news reporting in every area of the media–but they added in data from two "liberal" editorial pages and completely obscured that key finding. Beyond that, they made an utterly laughable choice of which newspapers might comprise a sensible "sample." But then, if you’re inclined to think that sixty days is a hundred, maybe you think the Post and the Times can stand in for all of the rest.

Wednesday: Pew fails to note the obvious reason for a drop-off in stories on Bush.

We’ve got your numbers right here: What did Pew learn about the Bush/Clinton coverage? We’ve got your numbers right here:

All those figures deal with news reporting. (Caveat: The Newsweek data include news reporting and opinion columns.) In all those areas, Bush came out ahead. Here are the data on editorials/op-eds at the Post and the Times, the only area where Clinton did better:

*Figures in this one section were deduced from Pew’s reported data. All other data are drawn directly from the Pew report.