R. Douglas Arnold sets an ambitious goal: “This book is the first large-scale study of how local media outlets cover members of Congress” (pp. 3–4). His ultimate success exemplifies how content analysis can illuminate a subject with empirical and systematic findings.

The book is a clinic in the application of the interdisciplinary research technique of content analysis. In general, the measurement stage of content analysis begins by choosing what media outlets to study. Next, the population of stories (Websites, or other cases) must be defined using objective criteria. Finally, the stories are coded based on a defined list of characteristics. Presented with a methodological challenge, Arnold describes and defends the choices that he makes.

A transparent methodology underlies his primary data set. Arnold chooses to study the local newspaper as the medium with the greatest potential for covering individual members of Congress. Dividing the nation's daily newspapers into categories based on circulation, he randomly samples each category to identify 25 local newspapers whose archives were available in the Data Times or Lexis-Nexis databases. After random sampling to identify one member of Congress for newspapers whose primary circulation area has multiple members, he defines a population of stories by a computerized search for the names of the 25 representatives in the 25 local newspapers between January 1, 1993, and November 8, 1994. After retrieving all 8,003 unique stories and opinion pieces mentioning the 25 representatives, he codes for 68 characteristics. Coding objective measures, such as whether the story was filed by a District of Columbia reporter, lays the groundwork for more subjective measures about emphasis and valence.

To increase confidence in his primary data set and to broaden the research, Arnold constructs three additional data sets. A second data set brings together the data from six newspapers having competing newspapers in the same city, with additional parallel data from six competitors. A third data set includes 61,084 article citations for a broader range of newspaper-representative pairs. Finally, a fourth data set links the amount of newspaper coverage of the representative in the locality with 675 of 1,795 cases in the 1994 National Election Studies data set.

The impressive range of data allows Arnold to draw some conclusions about accountability. He begins with the premise that a “rich informational environment increases the chances that citizens will have an evidentiary basis for determining whether they approve or disapprove of a representative's performance in office” (p. 12). His central argument, then, becomes that the best local newspapers do make an important contribution to holding members of Congress accountable for their performance in office.

In exploring accountability, Arnold offers a variety of perspectives. He begins by examining the volume of coverage of representatives. Using regression analysis, he explores how characteristics of the member of Congress and the local newspaper affect the quantity of newspaper coverage. Utilizing his primary data set of 25 representatives and 25 local newspapers, he then provides a systematic portrait of the content of news coverage. Next, he seeks to explain the reasons for differences in coverage of the same representative by looking at competing papers in the same city. Finally, he offers a “fledgling attempt” (p. 261) to demonstrate impact with data showing a positive relationship between citizen knowledge and the quantity of local newspaper coverage of the representative.

This exploration constitutes a significant contribution to our understanding of Congress and the news media. The book extends knowledge about local media coverage of members of Congress with systematic data about many newspaper-representative pairs over a sustained period of time. One encouraging finding is that congressional elections have substantially less horse race coverage than presidential elections. Arnold is less encouraged by coverage of policymaking beyond the roll-call vote: “The lack of regular coverage of hearings, markups, and coalition building makes it much easier for representatives to avoid the heavy lifting that is required to make things happen on Capitol Hill” (p. 254). The book reveals local news coverage as an overall advantage for incumbents but also as a potential counterbalance to incumbent advantage. In particular, the author finds that criticism of a member of Congress is much more likely in letters to the editor (40%) and editorial/op-ed columns (26%) than in regular news stories (6%).

Although the temporal backdrop is the end of an era, Congress, the Press, and Political Accountability has a timeless feel. It examines newspaper coverage of the 1993–94 Congress, after which 40 consecutive years of Democratic control of the chamber would end. Consistent with the goal of a more timeless work, Arnold chooses not to address the 1994 Republican takeover, except for the empirical finding that the Republican Contract with America was largely ignored in local news coverage of representatives. That I count only 10 of the 25 primary data set's representatives remaining in the House of Representatives or serving in the Senate (Richard Baker, Michael Crapo, Bob Filner, James Inhofe, Peter King, Jim Kolbe, Jon Kyl, Jim McDermott, John Spratt, Albert Wynn) when the 109th Congress began in January 2005 suggests that it was a wise decision not to dwell on substantive points related to individual members. Rather, the book's focus is on measuring concepts that are timeless, such as whether news coverage emphasizes policymaking, interacting with constituents, or electioneering. The timeless quality seems to guarantee that readers will be learning from this book well into the future.