Those who are troubled by the ubiquity of public opinion polls, particularly members of the press and public, often complain that polling has turned political leaders into followers, afraid to champion policies unless they are already broadly popular. Bill Clinton was often accused of conducting a poll-driven presidency, particularly after his party lost control of Congress. Similarly, George W. Bush's key political adviser, Karl Rove, is seen by many as the architect of a strategy in which nearly all of his boss's actions are taken with at least one eye on reelection.

These claims are, of course, exaggerated. President Clinton made a number of choices that were, if not clearly unpopular, at least not obviously “safe.” As for President Bush, the war with Iraq, regardless of one's feelings about its necessity, was unquestionably a triumph of opinion leadership. There were no polls showing serious public demand for toppling Saddam Hussein until the administration forced the issue onto the national agenda. Anecdotally, there is plenty of evidence that the claim that polling has turned the presidency into a “permanent campaign” is exaggerated at best.

Nevertheless, it is important to demonstrate such things empirically, and this is what Diane J. Heith attempts to do in her book. Heith's major contribution to this debate is her use of archival data from five administrations (Nixon through Bush I), as well as public documents and memoirs from those five and the Clinton presidency. These archival data enable the author to document every instance in which public opinion information shows up in “formal and informal memoranda, handwritten notes, pollster reports, and other written documentation” (p. 9). This, in turn, allows her to draw conclusions about how polling data are used, by whom, and to what effect. The result is an interesting look into the ways in which six presidents gathered and evaluated information about public opinion.

In her first chapter, Heith discusses the relationship between polling and theories of presidential leadership and also introduces her data set. In particular, she discusses the idea of the permanent campaign and distinguishes it from traditional notions of governing that involve consensus building and compromise. She lays out her argument that while “[t]he behavior and values of a campaign have indeed penetrated the White House, [they] have not overwhelmed or subsumed the governing process” (p. 8).

The second, third, and fourth chapters are devoted, respectively, to discussions of who is most likely to seek and use polling information, what sorts of data receive the greatest attention, and how presidents use polls to identify supporters and adversaries within the electorate. Heith argues that the White House generally attributes the highest value to information that can help advisors assess voters' reactions to various policy alternatives and sell preferred policies to the public. She notes that “[a]pproval ratings,” despite their prominence in the news media, “were not the presidential public opinion data of choice” (p. 57). (Of course, it is also possible that with presidential popularity figures so widely known and broadly available, it was simply unnecessary for the White House to generate them or discuss them in much detail.)

In her final chapters, Heith uses various case studies to demonstrate her point that while all modern presidents make extensive use of polls, they are “not abandoning traditional governing behavior in legislative battles in favor of a campaign approach” (p. 121). One exception to this rule, however, comes in those rare cases when presidents seek to avoid impeachment and removal from office. In such cases, and especially that of Bill Clinton, White House polling strategy becomes more consistent with the expectations of the permanent campaign model. As Heith correctly points out, surviving scandal has little in common with developing and passing a legislative program. During times of (possible) impeachment, presidents are, in effect, campaigning to remain in office: Single-minded attention is devoted to “winning,” and it is generally recognized that maintaining public support is essential to prevailing in the upcoming battle with Congress.

All in all, this book represents a worthy effort to bring some new evidence to bear on important questions of presidential governance. There are, however, a couple of potential quibbles that could be raised. First, the author's efforts to debunk the notion of the permanent campaign result in something of a straw-man argument. I am not aware of many political scientists (as opposed to pundits or voters) who actually believe that presidents have abdicated their duty to lead and to bargain with Congress in favor of slavish devotion to public opinion polls. It hardly comes as a surprise, therefore, when she concludes that most administrations, while they may carefully scrutinize polling data, generally continue to govern in the same ways that their predecessors did.

In addition, a significant amount of data is presented, and sometimes the forest seems lost in the author's effort to detail the trees. It is useful (though not shocking) to learn that different presidents handle public opinion data in different ways, but Heith might have done more to tie these findings together. I do not wish to exaggerate this point; there is some very good summary discussion. But at times, it was difficult to see how the author's data supported her broader argument.

Despite these concerns, Polling to Govern is a very useful addition to the literature. It should be read by students of the presidency and scholars concerned with American public opinion. It is written clearly and accessibly and should be of value to undergraduates, as well as graduate students and faculty.