A major aim of the burgeoning research on media and politics is to specify the empirical relationships between media content and individuals' political responses. For a variety of reasons, including the difficulty of learning which individuals have been exposed to which specific media content and the ever-present challenge of demonstrating causality, the answers have been elusive. These two well-written volumes take us several steps in the right direction.

In Echo Chamber, Kathleen Hall Jamieson and Joseph N. Cappella provide a rich textual analysis of what they regard as the Republican Party's vital allies in the media: conservative talk radio (Rush Limbaugh in particular); Fox News programs with Sean Hannity, Carl Cameron, and Brit Hume; and the Wall Street Journal's editorial page. Using content analysis, they argue that these media share similar lines of argument, which contrast sharply with those of the mainstream media, and that they define the mainstream media as being liberal, biased against conservatives, and therefore untrustworthy as information sources. The intent of these right-wing media, Jamieson and Cappella posit, is to insulate their audiences from contrary viewpoints, inoculate them against any mainstream sources they happen upon, and teach them how to argue with these sources—in short, to marginalize the mainstream media as well as Democrats and liberals for their conservative audience.

Jamieson and Cappella's discussion of cases is among the most interesting parts of their analysis. They suggest, for example, that in the controversy over former Senate Majority Leader Trent Lott's praise of Strom Thurmond, these conservative media sources worked actively to present frames that Republicans could use to defuse the crisis—such as the contention that the mainstream media used a double standard in criticizing Lott but not Democrats with a segregationist past—and to guide the crisis to an acceptable solution (in this case, Limbaugh's claim that Lott had gone too far in apologizing for his misjudgment and should therefore step down from his leadership position). Thus, they contend that the conservative media have been an essential part of the Republican Party's promotional structure since the 1990s, disseminating the Republican National Committee's framing of particular stories and “help[ing] vet candidates in Republican primaries for their loyalty to Reagan conservatism” (p. 239).

Jamieson and Cappella's causal argument—that exposure to conservative media produces attitude change consistent with the media content, rather than that people self-select into the conservative media audience because they already hold these attitudes—is perhaps the weakest part of their analysis (not surprisingly, given the difficulty of establishing causation). The results they present from a 1996 experimental study are not as clear-cut as one would hope, a point the authors acknowledge. They are on stronger empirical ground when they refer to these processes as being “mutually reinforcing spirals of effect and exposure” (p. 83). Their claim would be even more persuasive with greater attention to the falsifiability of their hypotheses: what evidence would be needed to show that conservatives and the Republican Party had not been guided by Limbaugh, Fox, and the Wall Street Journal in the Lott controversy, and is such evidence possible to obtain?

It would be easy to get the impression from Echo Chamber that the new conservative media pose a threat to American politics, with their use of ridicule, disparaging labels, the creation of enemies, and an emotionally charged vocabulary, from “death tax” to “partial birth abortion” (pp. 178–79). (The authors do, in fairness, make clear the benefits of the accountability provided by Limbaugh, Fox, and others.) Keep in mind, however, that mainstream political institutions make comparable efforts to maintain public support by creating enemies and defining contested terms in an emotional manner (see Murray Edelman in Constructing the Political Spectacle, 1988). These activities, in short, are at the heart of political behavior. Polarized politics, with activists on both sides increasingly divided in their views and mistrustful of one another, is not pleasant to witness, but it can enhance accountability in governance.

The authors do political scientists a major service by taking a systematic look at the modern conservative media. In a JSTOR search of political science journals, I found that articles having to do with Democrats outnumbered those having to do with Republicans by a ratio of 2.6 to 1 (49,226 to 19,204), and search references to “liberal” (57,093) far outnumbered references to “conservative” (38,083). There may be many reasonable explanations for this disparity, but it is vital for researchers to understand political movements of all varieties.

Tom Rosenstiel and colleagues examine another form of media accountability in We Interrupt This Newscast: How to Improve Local News and Win Ratings, Too, the report of an extensive five-year study of local news funded by the Pew Charitable Trusts through the Project for Excellence in Journalism. Although local TV news is an important source of political information, with a larger audience than that of national network or cable TV, local newscasts have experienced sharp declines in their ratings; during the past ten years, local TV newscasts have lost almost 20 percent of their viewers.

Using audience data and content analysis of news broadcasts on 154 local stations, the authors demonstrate that the typical “local” newscast does very little to reflect the community it serves. Local newscasts around the United States follow the same tired formula. They open with one or more sensational stories about crime, accidents, fires, and other public safety issues. At the end are one or more human interest stories that have been mentioned during the broadcast as teasers, designed to keep viewers from changing channels. The most important news of the day, featuring government, social issues, business, and international news, is sandwiched in between. These complex issues are chopped into stories of less than a minute in length, with fewer sources and much less detail than the sensational lead stories, and lack the background needed to make the issue comprehensible.

If local stations were sacrificing their democratic responsibilities in order to gain profit, it would be disturbing enough. But the authors show that the sacrifice seems to be in vain: these choices are bleeding the ratings of local TV newscasts. “Stations with higher-quality newscasts do better commercially than other stations even taking into account network affiliation, time slot, and the size and competitiveness of the market” (p. 29). The “quality” that “sells,” they find, is characterized by a comprehensive local focus, stories of significance (as opposed to unique crime stories, celebrity news, and human interest), active investigations, accuracy and fairness in presentation, the use of more information sources with appropriate expertise, and professional presentation (stories with coherence and explanation, as well as good production values). Sensationalism—gore and graphics beyond what's necessary to convey information—does not increase ratings.

There is good news and bad news in their findings. The meatier stories that increase viewership do not include long presentations of such vital matters as economic policy (nor do they include long stories at all; a minute and a half appears to be the audience's limit). Rather, a slight ratings advantage goes to health stories and those about political scandal or malfeasance. On the other hand, the “horse race” stories so common in campaign coverage did significantly less well in the audience ratings than did stories on candidate qualities and debates about issues (p. 145). And although 60 percent of stories about controversial topics presented only one side of the dispute (p. 114), these one-sided stories got somewhat lower ratings than did those presenting a mix of opinions. The authors conclude that in a changing media environment with rapidly multiplying sources of news, stories providing expert sources, context, and information as to how the event or issue affects the audience's lives are central to the success of local TV news.

The study has the great advantage of using Nielsen ratings from meters that automatically record when the TV is turned on and what channel is being received, supplemented by viewer logs, rather than being based on respondents' unreliable reports of what they've watched. But readers looking for a thorough empirical analysis will be disappointed. Much of the book's statistical analysis is presented in order to illustrate points, not to analyze findings in detail. The independent variables do not include some potentially important influences, such as the socio-economic, racial, and ethnic characteristics of the market.

In some ways, the study is an intriguing media analogue to Richard F. Fenno's Home Style (2002)—in this case, a look at local station managers' perceptions of their audience, just as Congress members hold perceptions of their constituencies. Although station owners expect very large profit margins (40 percent, which the authors say is four times higher than that of most American industries), the information about audience reaction on which they rely is generally of poor quality, often derived from consultants with little local expertise. Viewers seem to have “a much broader definition of news than the news professionals” (p. 89), resulting in an unmet need for news that goes beyond discussion of fires, police tape, and celebrities. This study shows how the need might be met effectively and profitably as well.