The funeral of Ronald Reagan marked a celebration not only of the president's political accomplishments but also of the idea that the core of presidential leadership is mastery of the bully pulpit. Published shortly after Reagan was laid to rest, Bill Clinton's autobiography also reflects our modern preoccupation with rhetorical leadership. Clinton credits many of his political victories—most notably fending off an impeachment charge—to the power of rhetorical appeals. He also attributes many of his failures to an inability to communicate effectively. George Edwards thinks that Reagan, Clinton, and the conventional wisdom they exemplify are just plain wrong. In a thorough and forcefully articulated study, Edwards argues that public opinion is never altered by presidential speech. Efforts to advance a president's political agenda through rhetorical appeals over the heads of Congress to the people are futile wastes of time and energy.

Edwards is not the first scholar to warn of the pitfalls of the rhetorical presidency, but he is the most unequivocal in his critique of it. Previous accounts of the limits of rhetorical leadership prompted the question: Under what conditions are presidents likely to succeed or fail? The notion of a “limit” implies that there is some possibility of success. Even those scholars who think successful rhetorical appeal to be very unlikely still concede that there are some conditions under which popular leadership is effective and necessary. Replace the word “limits” in the subtitle of this book with the word “futility” and one gets a better sense of the central claim. Edwards thinks that attempts to explain the conditions for rhetorical success need to be replaced by a different kind of question, such as why presidents continue to adopt strategies doomed to failure.

The crux of the argument is presented in Chapters 2 and 3. The author reviews every major presidential policy initiative and every major speech between 1981 and 2003, a period that includes the Reagan and Clinton years. He makes the good methodological point that the limits of the bully pulpit for all presidents can be revealed by its failure for the most rhetorically skilled presidents. He distinguishes between transformative leaders who persuade the public to change its mind and facilitative leaders who push the public in a direction toward which it is already headed. The argument of these chapters is that there have been no instances of transformative leadership. Presidents have rarely seen their poll results improve by 6 percentage points or more, either for approval of their job overall or for specific policies. In the relatively few instances in which there were significant opinion shifts after a speech, Edwards finds these due to factors other than the speech (for example, to events themselves to which the speech is a response, like the terrorist attack on the World Trade Center.) Presidents cannot fundamentally change public opinion. They cannot “reshape the contours of the political landscape.” At best, “they may endow the views of their supporters with structure and purpose and exploit opportunities in their environments to accomplish their joint goals” (p. 74). This is a provocative and important thesis that merits the attention of students of the presidency, journalists who cover the White House, and presidents themselves.

On Deaf Ears reads like a well-crafted and well-written legal brief. Almost all competing hypotheses and counterexamples are anticipated and discussed. However, like a lawyer's brief, the discussion is asymmetrical. Counterexamples are treated critically and invariably dismissed, while supportive evidence is always accepted at face value. Clinton's approval increased 10 percentage points after two speeches on Iraq in December 1998. Edwards attributes this result to the public's opposition to impeachment, rather than to the speech. Clinton's approval increased 7 points after his 1996 State of the Union address. Edwards mentions this but dismisses its significance without comment (p. 33). Support for Reagan's Strategic Defense Initiative increased significantly after his speeches, according to the Gallup data that Edwards uses for all of his cases, but he dismisses its significance because the White House's own poll showed some decline in support for SDI over a two-year period. But the White House poll shows support at a higher level than Gallup recorded in 1986, declining to a level in 1988 that is also higher than the level reported by Gallup (p. 58)! Apparent presidential successes are invariably treated as actual failures. On the other hand, apparent failures of presidents to move public opinion after a speech are never interrogated for hidden success. It is at least plausible, in some instances, that presidents halt or interrupt greater loss of public support than may have happened absent a speech.

Competing findings by other scholars are merely mentioned and dismissed, rather than refuted. For example, in a study of public opinion between 1949 and 1980, Lyn Ragsdale found that speech making positively affects the president's level of public support, and that its magnitude of influence exceeds that of positive events, military activity, consumer prices, and unemployment (American Political Science Review, December 1984). We are reminded of this study but not told why or where it falls short (p. 27).

Chapters 4 through 7 are devoted to explaining why presidents have such difficulty moving public opinion through speech. Attributes ascribed to the speaker, the message, and the audience are described to show how they serve to inhibit the president's effectiveness. Presidents continue to devote wasted energy to a rhetorical strategy because of the way they were politically schooled, through the primary-based selection system. Presidents also may “go public” for reasons other than affecting public opinion polls—to alter the national agenda, satisfy a constituency with symbolic benefits, prepare the public for a policy shift, or to alter elite debate. These objectives are offered as alternatives to the effort to move public opinion. They also raise the question of whether public opinion is always, or best, captured by public opinion polls. Taken together, might not these “alternatives” to seeking public support sometimes be components of a longer-term sophisticated strategy to shape public opinion and a political legacy?