| The Liberal Bias of Occasionally Noticing The Truth
As part of Random Lengths' coverage of the 10th anniversary of the 1992 riots, I wanted to take a deeper look at the root causes behind the continued conditions which remain largely unchanged. The result was my article on social dominance, a comprehensive theory explaining the systematic cohesion behind multiple facets of racism, sexism and ethnic prejudice. Most news media spend far too little time trying to illuminate why things happen. Too much "what" and too little "why" leaves people feeling overwhelmed, helpless, even desperate. That's not good journalism as I understand it.
Perhaps that's why I was genuinely puzzled when "Notable Quotables," a newsletter we get from the conservative Media Research Center, presented the following as an example of "liberal bias" in the media:
Probing Liberal "Root Causes"
"Today [South Central Los Angeles dentist Bill] Faulkner's office is back in that shopping center....But the root causes of the riot, he says, are still there....When it comes to the people, little has changed here in ten years. South Central still has a high unemployment rate and a high crime rate. Los Angeles remains deeply divided along racial lines with an enormous wealth gap between the rich and the poor." -- ABC's Judy Muller on South Central Los Angeles ten years after the riots, April 29 World News Tonight.
I found this passage as bland as custard. It did nothing to explore the "root causes," much less to look for the root causes behind the root causes -- as I tried to do by writing about social dominance. It just mentioned them -- nothing more. It said nothing about who or what was responsible for them, or how we might address them. "So, where was the bias?" I wondered. I decided to find out.
I talked with "Notable Quotables" co-editor Richard Noyes, who explained, "The idea behind this was... how the focus [was] on the root causes of the riots. As if there's a broader sociological phenomena that needs to be addressed moreso than individuals being responsible for their own behavior, their own conduct."
I still didn't get it. Muller wasn't excusing the rioters, and individual responsibility gets talked about every day, while the continued divide between South Central and Bel Air rarely gets the scrutiny it deserves. "But isn't there a need to look at root causes?" I asked. "That is one argument," Noyes responded.
Now I thought he was missing the point: "But, the idea of root causes is not necessarily opposed to the idea of individual responsibility, its not an either/or". I was right. He was missing the point. He readily agreed that it wasn't either/or, and then went right back to complaining, "But this story only looked at that one side of the coin". Noyes -- like most conservatives -- just doesn't seem to understand what "root causes" are. They aren't the other side of the coin, they are the reason you toss a coin in the first place -- a whole different level of explanation. If "root cause" explanations don't favor conservatives, they should blame reality -- not the media.
Conservative media critics, in particular, have a hard time understanding the very concept of "explanation". Social scientists, like physical scientists, use mathematics to model their theories and test their hypotheses -- as well as those of others. It's not enough to simply claim, as Noyes does, that "the mainstream press, which a lot of Americans get their news from, is overwhelming populated by people with liberal beliefs. Those beliefs bleed over into the coverage, and the conservative point of view is usually under-represented." You have to actually prove some sort of causation-and prove that it dominates over other factors that impact the news as well. That's how science works. That's how reality works.
The image of a liberal-dominated media doesn't square with what conservatives say when their guard is down. For example, way back in 1988, Adam Myerson, then-editor of the Heritage Foundation's Policy Review, wrote, "Journalism today is very different from what it was 10 to 20 years ago. Today, op-ed pages are dominated by conservatives....but this creates a problem.... If Bill Buckley were to come out of Yale today, nobody would pay much attention to him... there are probably hundreds of people with those ideas [and] they have already got syndicated columns."
Even before that, overall newspaper management was staunchly right-of-center. Only twice since 1932 have more newspapers endorsed a Democrat for President than a Republican. And a 1998 survey of Washington journalists found that most called themselves "centrists," as they lined up to the right of the public on a wide range of economic issues. For example, 59% of the public rated protecting Medicare and Social Security as a "top" or "single highest" priority, compared to 39% of journalists.
If you really believe that conservative viewpoints are under-represented, consider how Social Dominance has been neglected by the media (my article on it is the first I've seen in print), while The Bell Curve, which offered a conservative "root cause" explanation to the wealth gap -- i.e. blacks are inferior -- got massive media play despite the fact that social scientists ripped it to shreds. That's the problem the anecdotal approach that conservatives are so fiercely wedded to: there are anecdotes supporting both sides, but there's no sense of what they add up to without using scientific methods, without digging past mere appearances into actual patterns and their causes.
It may sound like a lot of hard work, time-consuming number-crunching which can be interpreted a dozen different ways, so it still doesn't prove anything. But it's not. At least not always. Consider the example of Bernard Goldberg's recent "liberal media"-bashing bestseller, Bias. It's loaded with anecdotes, and mighty light on anything resembling a scientifically testable hypothesis, much less evidence. But it did claim that "In the world of the Jennings and Brokaws and Rathers, conservatives are out of the mainstream and have to be identified. Liberals, on the other hand, are the mainstream and don't have to be identified."
Stanford researcher Geoff Nunberg, decided to test this, using a database of about 30 papers. He found the word "liberal" attached to five liberal politicians 4.21% of the time, compared to 3.2% for the "conservative" label attached to five conservative politicians. (A number of conservatives tried to refute him, using methods Nunberg found variously flawed. Even disregarding their flaws, none showed the sort of overwhelmingly disparity Goldberg alleged. All differences were far too small to be detected without counting.)
But does labeling a position necessarily indicate you disagree with it? Conservative media darling Andrew Sullivan thought so. Trying to counter criticism of Goldberg, he cited a statistical analysis showing that the NY Times used "right-wing extremist" at least six times more than "left-wing extremist," since 1996. "There are so many problems with this technique that it would take a whole book to explain them," wrote Bob Sommersby, of The Daily Howler. Easier to just run a simple experiment: with a 9-1 ratio over the same period the right-wing Washington Times tested out as even more liberal the NY Times.
Simple statistical analysis wrecks havoc with MRC's work as well. At year end they select groups of related "Notable Quotables" and send them to a panel of conservative judges to select their "Annual Awards for the Year's Worst Reporting". There were fourteen categories for 2001. Not be outdone by messieurs Sommersby and Nunberg, I did a sophisticated statistical analysis of my own: I counted the selections. Only 25 of 65 items (less than one every two weeks) could plausibly be called "reporting" -- news items, that is -- and 10 of those were a stretch, at best. Almost all the rest were opinions or analysis in various settings. (Exception: a cable broadcast of Helen Thomas introducing former President Clinton won a first place bias award.)
So "Worst Reporting" was a lie -- or as I put it diplomatically to Noyes, it was "confusing". He said I was "the first person who has expressed any confusion with that title". That's a good indication of how little scrutiny is given to claims of "liberal media" bias. No one stops to ask the simplest questions when such claims are made. The accusation alone is treated as its own proof.
Only slightly more sophistication was needed to discover that MRC's judges found less bias on network TV than on cable, and less in news than opinion pieces.
Average scores were higher for stories on cable (55.08), syndicated TV (52) and websites (48) than for magazines, network TV, newspapers, and local TV (all tightly bunched from 40.17 to 39.40) Only two of 14 first places went to network TV -- and one was in a category explicitly limited to network morning shows. Cable got 5, websites 2, magazines 2, local TV 2, syndicated TV ("The McLaughlin Group," no less!) got 1.
Non-news stories averaged 47.5 compared to 39.04 for news reports, and 37.0 for the subcategory of hard news, second-lowest behind the catch-all "other," with only two items. First places went to 12 non-news stories, no hard news stories, and only two soft news pieces -- one in a category limited to soft news.
Without MRC's blinders on, these figures speak loud and clear -- even without questioning the legitimacy of any supposed example of bias: they find relatively little bias on network TV news, where most of their wrath is directed. Typically, ABC Weekend Anchor Carole Simpson was cited three times -- and won a first place...for her online opinion columns, not her on-air reporting.
MRC claims that none of this matters. When I pointed out the lower bias scores for network TV, Noyes said, "Okay, well I think that probably misses the point. I think what, particularly with our judges, they do tend to focus on personalities. So the fact that Bryant Gumble up until last week was on broadcast, you ended up with a large number of votes that went into that because, you know, people recognized him as a figure who they regard as having a liberal ax to grind. It's not by type of media so much as it's by the personality of journalists and where they do tend to show up."
Call me old-fashioned, but focusing on personalities rather than the content of news stories sounds a lot more like a witch-hunt than a scientific search for truth. It makes perfect sense, of course, if your purpose is purely political -- to put people on the defensive, and drive them off the air. (Just ask Bill Mahr, who got First Place in "Damn Those Conservatives" and Second in "Blaming America First".) MRC claims it's difficult to look at content. As Noyes explains, "The underlying background of a lot of these people is liberal. To greater and lesser degrees, they keep it in check in their coverage or it bleeds through in their coverage, and the premise is that it biases their report."
Perhaps -- if you accept the right-wing view of who's a liberal, and ignore all the contrary evidence of conservative journalistic bias. But even then, there are other biases working as well: the dictates of bosses, shareholders, and advertisers just for starters, not to mention the pressure for ratings that drive coverage toward the sensational, and often frightening, and away from root causes. This helps explain why, for example, people continued thinking that crime was going up in the late 1990s, years after it had begun to decline. The job of witch-hunting is to point fingers at convenient targets, but the job of science -- of understanding root causes -- is much harder: It lies with tasks such as disentangling different influences and seeing which ones predominate. It's much easier to just tag journalists as "liberals" and then use everything they say anywhere to discredit their work on the job.
No single approach can tell us everything we need to know about media bias, but two recent studies done for Fairness and Accuracy Inreporting (FAIR) give important indications. The most recent study, conducted by Media Tenor, covered ABC, CBS and NBC network news in 2001. It found that 92% of all U.S. sources interviewed were white, 85% were male and 75% were Republican, where party affiliation was identifiable. Corporate representatives appeared about 35 times more often than union representatives, 7% versus 0.2%. I asked Noyes, "isn't this what really matters -- who gets used as a source, not what the reporter might think of them in private?"
But Noyes would have none of it. According to him, the more often people appear on TV, the more often they're getting attacked. If corporate representatives appeared 35 times more often than union representatives, it signals, "a lot more journalistic scrutiny of corporations, trying to bring corporate people into grill them, and give them a hard time". Pity the poor white male CEO. Network TV news, it turns out, is conspiring on behalf of Native American working women, whom it never puts onscreen.
Noyes was equally dismissive of another FAIR study, reviewing the Nexis database for major papers and broadcasting transcripts for 2001. It found over 25,823 citations for the 25 leading think tanks, of which 48% were conservative, 36% were centrist, and 16% were progressive or left-leaning. Suddenly, Noyes became quite concerned about mixing together news and opinion, "You have editorials mixed in with news product, you have transcripts, you have hearing testimony, you have a great number of things that would not classified as news," he complained.
But unlike MRC, FAIR never pretended it was only covering news. Also unlike MRC, FAIR isn't mixing in websites with network news. Nor is it picking and choosing, or judging items subjectively. It's presenting a comprehensive snapshot of all citations in the same media outlets. Think tanks want to get their views out on opinion pages and programs, just as much as they want to be in the news -- and FAIR was measuring their success. If liberals really dominate every aspect of journalism, the results would be predictable: an overwhelming dominance of liberal think tanks. Finding the opposite is strong evidence against MRC's claim. But not for Noyes -- and he doesn't even need to look at the data in order to reject it.
His master-stroke is to ask how FAIR categorizes the Brookings Insitute. "Centrist," I tell him. "I would consider Brookings a liberal think tank," he replies -- as if his opinion were a matter of fact. And that's that.
When I asked Michael Dolney, who did the study, what evidence he'd use, he quoted from his 1998 report, "Brookings has long had a centrist or center-right orientation. As far back as the mid-1980s, Fortune magazine (7/23/84) was approvingly noting that 'Brookings Tilts Right.' Current president Michael Armacost was undersecretary of state in the Reagan administration and President Bush's ambassador to Japan. Brookings' two most prominent analysts served in Republican administrations. Their most visible foreign policy expert, Richard Haass, is formerly of George Bush's National Security Council. Domestic political analyst Stephen Hess helped edit the Republican platform in 1976, and served in the U.S. delegation to the U.N. under Gerald Ford."
I had only mentioned "a President who was a Nixon Administration official" to Noyes, and he brushed me off, saying, "I think Nixon has a number of liberals in his Administrsation. He was the one who appointed Harry Blackmun to the Court, if you recall." I got his message: I've made up my mind, don't confuse me with the facts. So FAIR, Nunberg, Sommersby and I -- and anyone else with half a mind to -- can do all the statistical analysis they want. The MRC and it's legion of followers will wave us off with a sniff, and go right back to fondling their trinkets of "bias," their precious bits of glass mistaken for diamonds.
Don't believe me? Here's the worst, most biased piece of hard news reporting according to MRCs panel of judges:
"The U.S. is actually the least generous of the industrialized nations. In Sweden, a new mother gets 18 months of maternity and parental leave, and she gets 80 percent of her salary for the first year. Mother or father can take the parental leave any time until a child is eight. England gives 18 weeks maternity leave. For the first six weeks, a mother gets 90 percent of her salary from the government and $86 a week thereafter. German women get two months of fully paid leave after giving birth. The government and the company kick in, and either parent has the option of three full years in parental leave with some of their salary paid and their jobs protected." -- Peter Jennings, April 19 World News Tonight, following a story on a study showing more aggression in children who attend day care.
That's it. The crown jewel of liberal bias. A rare bit of comparison showing how much better families are treated by European welfare states. Me, I think it's conservative bias every time a conservative rails against the welfare state, and the media doesn't set the record straight by reminding us what a real welfare state looks like. Reality, you see, is strongly biased to the left. Every once in a while, someone in the media notices this. And that's when conservatives really start to howl. |
STATE BLOGS
