Significant scholarly work suggests mass media professionals largely set the voices in their coverage to mirror that of mainstream government debate. While it is a compelling assertion, this view is only intermittently supported by data. This paper suggests that such theories enjoy inconsistent support because they fail to consider the interaction between journalists' incentives and the political climate. An economic theory is advanced here that the political journalists' objectives to please their superiors, please themselves, limit negative feedback, and expend minimum energy, in other words to minimize their costs and maximize their benefits, lead them to rely on tactics that vary with the scope of consensus support for an issue. This theory is tested by analyzing newspaper coverage of four political issues debated in Congress.
In a typical year, U.S. presidents give 600 speeches. They issue 90 executive orders and take positions on as many as 500 pieces of congressional legislation. Each year the president puts forward 100,000 pages of regulatory changes. In a typical year, members of Congress introduce more than 7,000 pieces of legislation. Congressional committees collectively hold 3,000 meetings per year, while the text of remarks made in congressional floor debates fills more than 20,000 pages each year. Across the country, state legislatures and county and city commissions collectively consider more than 2.5 million ordinances and statutes every year. In an election year, upwards of one million candidates seek political offices in the United States ranging from dog catcher to president. From all this-and countless other governmental and political acts and events-a typical U.S newspaper has room for fewer than fifteen political stories per day, and the average network newscast contains no more than five minutes of political coverage.
It is a daunting task to filter the entire supply of political happenings into a manageable set of covered events. But, then, so too is the process of deciding how to cover those chosen events. From an uncountable set of possible angles, journalists must arrive at the one they will use to present the story at hand.
How do the media operate to bring us the political coverage we receive? This is, of course, a fundamental question for understanding not only the media, but our politics. This paper advances an economic theory of media coverage of politics based on understanding the basic motivations of journalists and the importance of political consensus to their work. The theory is tested using a sample of newspaper coverage of several issues debated in Congress.
The Indexing Theory
Among the most prominent efforts to advance a theory of political media coverage is Lance Bennett's indexing theory.1 Bennett began with the simple notion that the mass media look to the government for most political information they report.2 In the process, he argued, mass media professionals-from reporters to publishers-index the voices in their coverage to that of mainstream government debate. In other words, political viewpoints found predominantly outside of the U.S. government or those found at the fringes of the dominant U.S. political parties are likely to be ignored or minimized.
Bennett presents indexing as a collective outcome. Indexing is not an openly taught method nor an articulated objective, but the consequence of practical decisions made by journalists and their superiors. Indeed, Cook writes of indexing as a consequence of journalists' strategic behavior: "Newspersons are less preoccupied with reflecting a legitimized elite than with generating a predictable stream of nonfiction stories about the world."3
Bennett acknowledged that certain areas of controversy (such as foreign affairs, trade) would likely produce a stronger adherence to government voices. But, he asserted that his work offered "a general theory of the press and the state in the United States. …