Media bias aids Democratic candidates "about 8 to 10 percentage points," in a typical election, a professor of political science at UCLA , Tim Groseclose, claims in his new book, Left Turn: How Liberal Media Bias Distorts the American Mind.
The book cites a series of experiments and research conducted by the author and others. In one, researchers looked at markets in which Fox News was available in 2000 and markets in which it wasn't. "George W. Bush's vote share was 0.43 percent higher in Fox markets than it was in non-Fox markets," the book reports. In another experiment, Yale researchers bought Virginia residents gift subscriptions to either the Washington Times or the Washington Post. The ones who were randomly assigned to the more left-leaning Post voted for the Democratic gubernatorial candidate 3.8 percentage points more than those who were randomly assigned the more right-leaning Times.
Other research analyzes the text of newspaper articles or television broadcasts, focusing on things such as how often sources from conservative or liberal think tanks are quoted, or whether "partial birth abortion" is called by that term or something else.
The result is a wonderfully entertaining and edifying little book, full of sparkling examples. One of my favorites is from January 31, 2004, after the Commerce Department reported 4% annualized GDP growth for the fourth quarter of 2003. "Economy Remained Strong in 4th Quarter," was the New York Times headline. "GDP Growth Disappoints," was the Chicago Tribune's headline. The author cites a study that for the same piece of news, "major U.S. newspapers are 20 to 40 percent more likely to report a negative headline if the administration is Republicans than if it is Democratic."
And that's just the research related to the content and effect of the coverage. Research about the political leanings of the journalists themselves is another area to explore: "in a typical presidential election Washington correspondents vote about 93-7 for the Democrat."
As a remedy to the bias, Professor Groseclose calls for journalists to be less secretive and more transparent about their own political views, so that readers or viewers are not deceived.
My main quarrel with this book is its design. The type is too small, there's too little space between lines, and the subheadings and tables combine with Professor Groseclose's occasional lapses into political science jargon ("media lambda," "media mu") to make what's really a brisk and accessible book appear far more impenetrable than it actually is.
Disclosures: The publisher sent me a review copy. As usual, if you buy the book from the link above this site gets a share of the revenues.