You Mean They Actually Research This Stuff?

For all the blather about Media Bias (to which I've contributed more than my fair share), it's worth pointing out that academics have actually researched bits of this question in great detail. Three interesting papers from 2004:

In March, MIT political science professors James Snyder, Stephen Ansolabehere and Rebecca Lessem published a study [PDF] of newspaper endorsements from 1940-2002. Conclusion:

First, newspapers have shifted from strongly favoring Republicans in the 1940s and 1950s, to dividing their editorial endorsements roughly equally between the parties. Today, Democratic candidates are about 10 percent more likely to receive an endorsement than Republican candidates. Second, newspaper editorials have come to favor heavily those already in office. Incumbents today receive the endorsement about 90 percent of the time. In the 1940s, incumbents received endorsements only about 60 percent of the time. The consequence of this shift, we estimate, is to increase incumbents' vote margins, on average, .2 to 1 percentage points.

In January, Snyder and Ansolabehere teamed up with Stanford's Erik Snowberg to look at [PDF] media coverage of campaign contributions from 1996-2000.

The average figures reported in newspapers exceed the analogous figures from the FEC by as much as eight fold. Press reports also focus excessively on corporate contributions and soft money, rather than on the more common types of donors—individual—and types of contributions—hard money. We further find that these biases are reflected in public perceptions of money in elections. Survey respondents overstate the amount of money raised and the share from different groups by roughly the amount found in newspapers, and better educated people (those most likely to read newspapers) showed the greatest discrepancy between their beliefs and the facts.

Italics mine. Also, the American Enterprise Institute's Kevin Hassett and John Lott compared coverage of similar economic news during different presidencies, and concluded that "American newspapers tend to give more positive news coverage to the same economic news when Democrats are in the Presidency than for Republicans."