FEAR AND FAVOR.

Just before last fall's Presidential election, Bill Keller, the executive editor of the Times, and Philip Taubman, the paper's Washington bureau chief, went on the road to inspect the candidates' campaigns. In Florida, on October 22nd, they arranged to have drinks with Karl Rove, the White House's chief political strategist, and Dan Bartlett, its head of communications. It was supposed to be a friendly get-together, and that's how it went for the first few minutes, until Keller asked Rove what he thought of the Times' coverage. It's the sort of question that editors often ask important people, in the same spirit that a politician asks, "How'm I doing?," usually hoping for an answer somewhere in the lower-middle range of politeness and candor. But Rove, Keller told me not long ago, "pounded on us for two cocktails' worth of conversation." Saying what? "It was three kinds of things," Keller explained. "It was Bush accomplishments we had ignored, flaws in the Kerry record that we had put inside the paper, and a number of pieces we had done looking hard at the Bush record. In their view, that all amounted to arming the Kerry campaign."

Keller and I were talking in his office in the Times newsroom at nine one morning, a moment when most newspaper offices are empty and expectantly quiet, like a theatre a couple of hours before the curtain. Keller took his time describing the conversation, to suggest that he wasn't dismissing the criticisms out of hand. "Your initial reaction, especially in someone as ferocious as Rove, is to drop into a defensive crouch," he said. "But I try not to do that. I listened, with a fair measure of skepticism, because a lot of it is calculated. But there was some genuineness to it. He went through a long litany of complaints. I do think he was channelling a feeling about the New York Times that's out there in the land, that we should be concerned about, or at least aware of."

One item that particularly drew Rove's ire was a Times front-page story, by Ford Fessenden, which appeared on September 26th, under the headline "a big increase of new voters in swing states." As Keller remembered it later, in an e-mail message to me, Rove "fired off complaints like a Gatling gun, some specific, some generic, some about specific writers, some about specific elements of specific stories." When I spoke to Rove about his conversation with Keller, it was obvious that, to his mind, the September 26th story was No. 1 among the Times' journalistic misdeeds during the campaign. The story left the impression that the Democrats' organization was vastly superior to the Republicans', especially in Florida and Ohio. Getting out the G.O.P. vote in those two states had for several years been one of Rove's main projects, and he spoke about the article in roughly the same tone as a writer discussing a bad review of his magnum opus. He gave me a highly detailed, twelve-point critique, and then, in the interest of conciseness, he boiled down the twelve points to two or three.

According to Rove, the Republicans, in their organizing, had probably far surpassed the Democrats in all of the swing states except Pennsylvania and maybe New Mexico. They had certainly done so in Florida, and arguably in Ohio. The Times story had generally relied on Democrats and groups affiliated with the Democratic Party for its information, and had got only pro-forma responses from Republicans. Fessenden had gauged new-voter registration by comparing figures for the first seven months of 2000 and the same period in 2004; framing the data that way favored the Democrats, because their organizing effort hadn't begun seriously until 2004. The Republicans, Rove said, had been organizing since 2001, but a comparison of only the year 2000 with the year 2004 omitted the progress that the Party had made during 2002 and 2003.

Another technique the Times had used was surveying new registrants in the most heavily Democratic and Republican Zip Codes in Florida and Ohio. But this, Rove said, was unnecessary in Florida, where voters register by party, and misleading in Ohio, where the Republicans were finding most of their new voters in precincts (not Zip Codes, which aren't political boundaries) that had not voted heavily Republican in the past. Rove felt that the Times had allowed itself to be fed by the Democratic organizations. Why did it seem as if so many sloppy errors in the Times' political coverage favored the Democrats? (Fessenden, when I told him of Rove's complaints, insisted that, ...