The old litmus tests that sort journalists into left and right don't work anymore.

Whenever conservatives talk to liberals about press bias—or vice versa—they talk right past one another. Both factions seem to work backward from their conclusions to the evidence and damn what the other side says. For a prime example, see the "Are the Media Liberal?" debate in National Review Online from last week. In it, conservative L. Brent Bozell III of the Media Research Center spars with lefty journalist Eric Alterman, author of What Liberal Media?

In his first response to Alterman, Bozell calls liberal media bias "obvious," "documented," and "proven" and cites a "national survey of the Washington-based media commissioned by the Gannett media organization" to demonstrate the press corps' essential liberality. (Actually, the survey was commissioned by the Freedom Forum foundation, which grew out of a foundation started by publisher Frank E. Gannett but is completely independent from the Gannett news corporation.) Bozell writes that the study found that:

in 1992, by 89-7 percent, [press members surveyed] voted for Bill Clinton over George Bush; that by 50-14 percent they see themselves as Democrats over Republicans; and that while 61 percent describe themselves as liberal, only two percent dare call themselves "conservative." ...

Such lopsided numbers would turn anybody's head. But how accurate are they? Not very. Reporter Robert Parry exposed the survey's weaknesses in a 1997 piece: The polling group that conducted the survey sent its questionnaires to 323 journalists covering some aspect of Congress. Only 139 completed surveys came back, and nine of them left the question about their presidential vote blank.

While you might reap accurate results from a Vulcan mind probe of just 130 members of the Washington press corps, you'd want to make sure the 130 surveyed were the right 130. But that's not the case with this survey. Only 60 of the 323 questionnaires went to journalists at the elite organizations that set the news agenda: the New York Times, Washington Post, Los Angeles Times, USA Today, Wall Street Journal, CBS, NBC, ABC, CNN, PBS, National Public Radio, Time, Newsweek, U.S. News, the Associated Press, and Reuters. Instead, the pollster told Parry, most of the surveys were sent to Washington staffers at regional newspapers (the Boston Globe, Denver Post, Dallas Morning News) or at the chain wire services (Knight-Ridder, Newhouse), with a quarter of the questionnaires going to pipsqueak pubs like Indian Country Today, Hill Rag, Washington Citizen, and Government Standard.

The survey guaranteed its respondents anonymity, so nobody knows who, exactly, returned surveys. But we can guess: Self-important big-shots surely round-filed the nosy Q & A, and flattered small-fry probably obliged. I'd be astonished if more than 10 of the 60 elite journos surveyed bothered filling out the form.

What's more, Bozell and Co. ignore the changing face of politics when they attempt to prove the liberality of the press corps by rounding up Clinton voters. Clinton ran as a controversial (within his own party) centrist—not a Carter, Mondale, or Dukakis liberal. He also governed from that slice of the political spectrum. Conclusion? Case most definitely not closed.

Alterman matches Bozell's lameness by "disproving" liberal bias with quotations from leading Republicans and conservatives—Rich Bond, James Baker, Pat Buchanan, and Bill Kristol. These figures either confess that they "work the ref" (the press) in hopes of winning favorable coverage for conservatives or concede that press bias really isn't that big a deal. Bozell zings Alterman for ignoring the content-analysis work he and other conservatives have done and for plucking a quotation here and a quotation there from the vastness of Lexis-Nexis to make his point.

The ongoing sermons about media bias from right and left stink mostly because they rely on the ideological frameworks constructed by the bias-hunters of the early '70s. In the old days, it was easy to sort Washington journalists into left and right by lining them up against the wall and giving them a dozen litmus test issues—for or against, say, reproductive rights; gay rights; civil liberties; gun control; national health care; school prayer; capital punishment; arms control; free trade; welfare; crime; civil rights; the drug war; corporate power; the environment; porn; and deficit spending. Anybody who took the liberal line 75 percent of the time qualified as a liberal.

But the old litmus paper has lost its magic. While the media-bias interlocutors' positions remain fixed—have any political views at the Nation changed in the last half-century?—the views of Washington journalists have shifted, as have those of the public. The journalists-formerly-known-as-liberals (TJFKNAL) are no longer reflexively against free trade, the drug war, and corporations, thanks, in part, to the centrist teachings of Bill Clinton. And they're no longer reflexively for deficit spending, welfare, and porn. Likewise, traditional conservative views about the environment, civil liberties, the deficit, arms control, and health care have morphed, allowing President Bush unilaterally to cut the nuclear arsenal, propose new drug subsidies for the codgers, and budget for deeper deficits.