July 5 (Bloomberg) — It's called the Fourth of July but it felt like Memorial Day to many journalists.
They spent yesterday mourning Rupert Murdoch's bid to snap up the Wall Street Journal. The prospect of Murdoch in charge had editors fretting that the media lord will impose his right-wing political views upon the knights of the Journal's newsroom.
The grieving goes beyond Dow Jones, the Journal's parent, to other industry buyouts. News hands believe that the consolidation must lead to a tragic reduction in quality and freedom. After all, as far back as 1945 the Supreme Court opined that "right conclusions are more likely to be gathered from a multitude of tongues." Everyone knows editorial pages shift, but the fear is that owners also will fiddle with the news columns.
This is a story that touches just about every journalist, pro-Murdoch, con-, or confused. I worked for Dow Jones for almost two decades, and my book publisher is HarperCollins, owned by Murdoch's News Corp. But do the data about bosses' influence match the emotion?
Not really, if you believe two University of Chicago economists, Matthew Gentzkow and Jesse Shapiro. Like many a newsman, the pair decided that they weren't interested in editorial pages and looked only at news. In a study for the National Bureau of Economic Research, they compared owners' politics to the political content of their papers' coverage of news.
They found that news coverage does indeed parallel the views of someone, but that someone isn't the boss. It's the reader. What's more, the authors found, owner diversity isn't a necessary condition for media diversity. The same owner often has newspapers with different slants in different places.
The Gentzkow/Shapiro method features many steps, but they are as straightforward as an old editor's ruler. The pair started by looking at donations made by media companies and executives to political parties or individual Democrats and Republicans. They couldn't get all data on all owners, but they were at least able to develop a standard. The folks at the top of Gannett Co. or MediaNews Group Inc., which publishes the Denver Post, were entirely Republican by this measure. The Seattle Times Co. management, by contrast, was Democratic.
Next, the authors measured the politics of the newspaper's customers by looking at voting patterns in counties where the papers are sold.
But it is when it came to determining political leanings in the news reports that they got truly ingenious.
First, they studied patterns in the remarks of members of Congress to discern which buzz words go with which party. Reviewing the appearance of a thousand key phrases in news databases and the 2005 Congressional Record, they found Democrats favor such combos as "tax break," "private account" and "war in Iraq." Republican lawmakers by contrast prefer the phrases "tax relief," "personal account," and "war on terror." Another Democrat favorite was "Corporation for Public Broadcasting," while Republicans liked "Circuit Court of Appeals."
Some of the differences were extreme. The Congressional Record shows Democrats said "estate tax" a total of 195 times, and "death tax" only 35 times. For Republicans, the rate for "estate tax" was 46, while they uttered the phrase "death tax" 365 times.
Using the appearance rates of the same phrases in news columns to gauge political slant, the authors then looked to see which papers lined up with which party. "We find," they wrote, "that the average newspaper's language is similar to that of a left-of-center member of Congress." No shock there. Even before the days of anti-Bush rage, newsroom Republicans were so rare that they could barely fill a lunch table in the cafeteria.
But there was a surprise. The politics of the papers didn't necessarily line up with the politics of their owners. They lined up, rather, with voting patterns in the zip codes that the papers served. This was true even for Murdoch's New York Post, whose news coverage turned out to be only an increment more "Republican" than that of the liberal New York Times. News article word choice, Gentzkow told me, "fits the taste of the readership in those city markets."
What about when the boss changes? Genztkow and Shapiro followed three newspapers that changed ownership during the period of their study, 2000-2005: the Los Angeles Times, the papers Thomson Corp. sold to Gannett, and 16 daily newspapers that Lee Enterprises Inc. acquired from Howard Newspapers.
In none of the cases was the change in political leaning of the paper statistically significant. Most interesting of all, it turned out that individual newspapers' politics might differ according to the politics of their audience, even when the newspapers have the same owner.
According to Locale
As the authors point out, the New York Times Co. has the Times in New York, a heavily Democratic city, but also a paper in Spartanburg, South Carolina, which is 67 percent Republican. Those papers' politics varied according to their locale, not Arthur Sulzberger Jr.'s mood.
One can challenge the thesis. On the chart of newsroom slant, the Washington Times sits way over on the right. But the zip codes that the paper serves are fairly Democratic. Owners sometimes do matter more than the local consumer.
Still, the evidence suggests that papers go with readers, not press barons. It is the guy in the next beach chair who is the true boss. That realization may dispel some fears, at least enough to make the weekend feel like a real holiday.
(Amity Shlaes, a visiting senior fellow at the Council on Foreign Relations, is a Bloomberg News columnist. The opinions expressed are her own.)
© Copyright 2007 Bloomberg
Available for order: